Efficacy of Alogliptin and Pioglitazone in Subjects With Type 2 Diabetes Mellitus: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Efficacy of Alogliptin and Pioglitazone in Subjects With Type 2 Diabetes Mellitus (NCT00432276), a randomized phase 3 study. The primary outcome is Change From Baseline in Glycosylated Hemoglobin (HbA1c).

2. Clinical Question

The clinical question is therefore a comparison of treatment strategies for people inadequately controlled on the stated combination at entry. One strategy retained pioglitazone 30 mg and added alogliptin; the other increased pioglitazone to 45 mg while using matching placebo tablets. Metformin was continued in both listed regimens at each participant’s maximum tolerated dose.

The registered secondary outcomes extend the question beyond average HbA1c change. They include change over time, attainment of specified HbA1c thresholds, and decreases of specified magnitudes. These are different ways to describe glycemic outcomes: a mean change summarizes the group average, while threshold and decrease outcomes describe the proportion meeting a stated criterion. The record lists these outcomes but does not provide their numerical results in the information.

3. Trial Design

The registry describes a Phase 3, randomized, double-blind, parallel-group study with quadruple masking. It lists an enrollment of 803 and identifies Takeda as the sponsor. The trial status is completed. The recorded start is 2007-01, and primary completion is 2009-05.

Randomization assigns participants to treatment groups by chance, which is intended to reduce systematic differences between groups at entry. In a parallel design, participants remain assigned to their group rather than crossing over between the listed regimens. Masking is intended to limit the influence of knowledge about treatment assignment on participants’ experiences, clinical management, or outcome assessment. The registry calls the masking quadruple, but does not specify in the information which roles were masked or how masking was maintained.

The two listed regimens share ongoing metformin treatment and differ in whether alogliptin is added while pioglitazone remains at 30 mg, or whether pioglitazone is increased to 45 mg with alogliptin-matching placebo. Both descriptions specify oral, once-daily study tablets and treatment for up to 52 weeks. The registry’s brief summary frames the study as a comparison of adding alogliptin to ongoing pioglitazone and metformin against the alternative pioglitazone regimen.

The primary endpoint is described at baseline and Weeks 26 and 52. The record’s primary-outcome population is the per-protocol set, not simply everyone randomized: it requires receipt of at least one dose of double-blind medication and absence of major protocol violations. Last observation carried forward was used for imputation in this population. The details of the randomization procedure, allocation concealment, site structure beyond the multicenter description in the official title, and any departures from the planned design are not reported in the registry record.

Completed status and enrollment describe trial conduct, not the size or certainty of a treatment effect. Without the numerical HbA1c results, the registry information does not establish the direction or magnitude of the primary comparison. The study design sets up a randomized comparison; the outcome estimates are needed to interpret what the comparison observed.

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: Participants with a diagnosis of type 2 diabetes who were experiencing inadequate glycemic control on their existing treatment regimen of metformin HCl plus pioglitazone were randomized in a 1:1 ratio to 1 of 2 treatment arms: addition of alogliptin 25 mg versus titration of pioglitazone 30 mg to 45 mg.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Per-protocol set included all randomized patients who received at least 1 dose of double-blind study medication and who had no major protocol violations. Last observation carried forward (LOCF) imputation was utilized.

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.

6. Primary Endpoint

The registry lists the primary outcome as Change From Baseline in Glycosylated Hemoglobin (HbA1c), measured over Baseline and Weeks 26 and 52..

The primary endpoint is change from baseline in glycosylated hemoglobin, or HbA1c, at Week 26 and Week 52. The record defines HbA1c as glucose bound to hemoglobin, expressed as a percentage of the absolute maximum that can be bound. It describes the reported outcome parameter as a least-squares mean and its dispersion as a standard error.

HbA1c is used as a continuous measure of glycemic status. A change score expresses how the measure at follow-up differs from the participant’s baseline value. Summarizing change across a treatment group can describe the average pattern while accounting for features of the analysis model, although the exact model used is not reported in the registry record. A least-squares mean is generally a model-based group mean; it should not automatically be read as an ordinary arithmetic average unless the analysis details support that interpretation.

As general statistical context, a standard error describes the precision of an estimated mean under the assumptions of the analysis. It is not the spread of individual participant measurements. A smaller standard error generally indicates a more precisely estimated group mean, but interpretation also depends on the estimate itself, the analysis model, and the comparison of interest. A confidence interval can convey a range of values compatible with the data under a specified method, but a confidence interval for this endpoint is not reported in the registry information.

HbA1c change is one summary of outcome, not a complete account of every participant’s response or experience. The record also lists threshold-based secondary outcomes, but those do not replace the primary continuous measure. Numerical primary-outcome values are not reported in the registry record, so the size and direction of HbA1c change cannot be characterized from these details.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry describes a continuous primary outcome summarized using least-squares means, with standard errors as the dispersion measure. The primary measurement is change in HbA1c from baseline at Week 26 and Week 52. The statistical model used to produce the least-squares means is not reported in the registry record, so the exact adjustments and assumptions underlying those estimates cannot be specified.

What does a least-squares mean represent?

As general statistical context, a least-squares mean is an estimated mean from a statistical model. It may be adjusted for variables or design features included in that model. This can be useful when an analysis seeks to compare groups while accounting for prespecified factors, but the meaning depends on the actual model. Since the model terms are not reported in the registry record, it is not possible to tell which factors, if any, were included or how the estimates relate to unadjusted group averages.

The endpoint is a change score, so it uses a participant’s baseline measurement as a reference for follow-up. Change outcomes can help focus on within-participant movement over time, but they still require an appropriate between-group comparison to answer the randomized treatment question. A change in one group alone does not establish that its regimen differs from the comparator. The record does not provide the numerical changes or a single comparison across the stated time points.

How should standard errors be read?

As general statistical context, a standard error quantifies the sampling uncertainty of an estimated quantity under the chosen model. It is distinct from a standard deviation, which describes variation among observations. Standard errors are often used in constructing confidence intervals or test statistics, but neither a confidence interval nor a test result for the primary endpoint is reported in the registry information. A standard error also does not measure bias from missing data, incorrect model assumptions, or other design limitations.

Randomization supports a fair comparison by assigning treatment independently of participants’ measured and unmeasured characteristics in expectation. Masking can reduce some forms of bias related to knowledge of assignment. These design features do not remove the need to examine outcome definitions, follow-up completeness, analysis populations, and statistical methods. The record states randomized allocation and quadruple masking, but does not report the implementation details needed to assess these procedures further.

What does the analysis population mean?

The primary population is described as per-protocol. It includes all randomized patients who received at least one dose of double-blind study medication and had no major protocol violations. This is a population definition based on post-randomization treatment receipt and protocol adherence. As general statistical context, restricting analysis after randomization can affect the balance created by random assignment, because eligibility for the analysis may depend on events occurring after assignment. That does not establish that bias occurred in this trial; it identifies why readers often compare a per-protocol analysis with an analysis that retains all randomized participants. Such an additional numerical analysis is not reported in the registry record.

Last observation carried forward, or LOCF, is the stated imputation method. As general statistical context, LOCF substitutes a participant’s last observed outcome for a later missing value. This produces a value for analysis but assumes that the last recorded measurement is an adequate stand-in for the unobserved follow-up measurement. That assumption may be inappropriate when the outcome could change after the last observation. LOCF does not describe why data are missing, and it does not automatically account for uncertainty about the missing values. The registry does not report the amount or pattern of missing data or sensitivity analyses that test alternative assumptions.

How should the multiple outcomes be understood?

The registry lists 41 outcome measures and 26 analyses. The primary outcome itself is described at more than one follow-up time, and secondary outcomes include change over time, HbA1c thresholds, and specified decreases in HbA1c. As general statistical context, examining multiple outcomes, time points, or comparisons can increase the chance that at least one result appears noteworthy by chance if each is assessed separately. Methods such as prespecifying a primary comparison or adjusting for multiple testing can address this concern, but the multiplicity strategy is not reported in the registry record.

The registry’s statement that the primary outcome has several categories or time points and is not reduced to one comparison is important. It means a single summary cannot be assumed to represent all primary-outcome results. The analysis plan for selecting or interpreting results across the time points is not reported. The outcome list alone does not reveal which estimates were emphasized, whether tests were hierarchical, or whether any adjustment was made.

What do the planning assumptions mean?

The planning illustration is described as using a two-sided significance level of 0.05, equal allocation between arms, and a normal approximation, with the effect to detect defined as a stated fraction of the observed effect. As general statistical context, such calculations illustrate how a hypothetical future design might relate its sample size to an assumed effect and error threshold. They are not observed evidence about the treatment comparison. The record does not provide a trial-specific sample-size justification or numerical planning outputs in the information. Any illustration based on an observed effect is not observed power, and should not be interpreted as proof that the completed trial was adequately or inadequately powered.

These methodological details allow readers to identify important features of the analysis without filling in unreported choices. The numerical estimates, model specification, multiplicity approach, missing-data pattern, and sensitivity analyses are necessary for a fuller statistical assessment and are not reported in the registry record.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists 41 outcome measures and 26 analyses. Its primary outcome covers change from baseline at Week 26 and Week 52, and the record states that the endpoint has several categories or time points rather than one reduced comparison. Secondary outcomes include change over time, HbA1c thresholds, and decreases of specified magnitudes. These multiple outcomes and time points create several possible ways to describe the results.

As general statistical context, when many outcomes or comparisons are examined, interpreting an isolated result can be difficult unless the analysis plan clarifies which comparisons were primary and how multiplicity was addressed. The registry record does not report a multiplicity adjustment, testing hierarchy, or other strategy for handling multiple comparisons. It is therefore not possible to determine from the reported details how the risk of chance findings across the outcome set was managed.

For the primary endpoint, the record identifies the per-protocol population and states that LOCF imputation was used. The population includes randomized patients who received at least one dose of double-blind medication and had no major protocol violations. The record does not report the number or pattern of missing measurements, reasons for missingness, or sensitivity analyses that assess alternative assumptions. As general context, missing-data methods rely on assumptions about unobserved outcomes, and different assumptions can affect estimates.

The information does not report whether an intention-to-treat analysis or other analysis population was used for the primary comparison. Nor does it report the detailed statistical analysis plan or deviations from that plan. These omissions limit assessment of how robust the results are to analysis-population choices, missing-data assumptions, and multiplicity. They do not establish that any particular bias or analytic problem occurred.

11. Safety Results

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

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: acute myocardial infarction (0.5%), cataract (0.5%), fall (0.5%), hypotension (0.5%), non-cardiac chest pain (0.5%).

The registry reports serious adverse event counts for each listed treatment group. In the alogliptin 25 mg, pioglitazone 30 mg, and metformin group, 20 participants were affected among 404 at risk. In the pioglitazone 45 mg and metformin group, 20 participants were affected among 399 at risk. For other adverse events, the record reports 116 among 404 at risk in the former group and 104 among 399 at risk in the latter group.

These figures are counts of affected participants alongside the registry’s at-risk counts. They are descriptive safety information, not an adjusted comparison. No risk difference, ratio, or statistical test is reported in the information, and none should be inferred by comparing the counts alone. Differences in exposure, follow-up, event definitions, or ascertainment can matter when interpreting adverse-event summaries; the relevant detailed methods are not reported in the registry record.

The listed serious-event terms are acute myocardial infarction, cataract, fall, hypotension, non-cardiac chest pain, and osteoarthritis, each reported at 0.5%. The record does not give group-specific counts for these terms in the information. Death counts are not reported in the registry record. Adverse-event counts should be read as outcomes observed and recorded during the study, not as proof that a treatment caused an event or as a complete measure of safety.

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

The registry record establishes that the study was designed to compare adding alogliptin to pioglitazone and metformin with increasing pioglitazone while continuing metformin. It also identifies the primary outcome as change in HbA1c from baseline at Week 26 and Week 52, and gives the analysis population and LOCF approach. These details explain the intended comparison and some of the choices used to summarize the endpoint.

They do not establish the primary efficacy result. The numerical HbA1c estimates, their standard errors, and a treatment contrast are not reported in the information. Without those values, the direction and magnitude of the observed comparison, and the precision with which it was estimated, are not reported in the registry record. No conclusion about whether one regimen produced a larger or smaller HbA1c change can be drawn from the design description alone.

The safety figures are also limited in what they can establish. They report serious and other adverse-event counts in each group, but do not provide a statistical comparison or a basis for attributing events to a particular regimen. The serious-event terms are listed with percentages, while death counts are not reported. These summaries should be interpreted as registry-reported event information, not as a complete account of all possible safety differences.

As general statistical context, a treatment comparison is more informative when its estimate is considered alongside an uncertainty interval, the analysis population, and assumptions about missing observations and multiple testing. The record describes a per-protocol population and LOCF but does not report missing-data patterns, sensitivity analyses, the detailed model, or the multiplicity strategy. Those gaps prevent a full assessment of how stable an estimate might be under alternative reasonable analytic choices.

The sample-size planning illustration, if considered, is a planning exercise based on stated assumptions and an effect fraction. It is not observed power and does not substitute for the trial’s actual results. The record does not report a trial-specific sample-size justification in the information. The appropriate interpretation is therefore bounded: the design and outcome definitions are reported, while key numerical efficacy results and several analysis details are not reported in the registry record.

13. Why This Trial Matters Statistically

This trial illustrates why a registry entry needs to be read as both a design description and a record of reported results. The randomized, masked, parallel design defines the intended comparison between two regimens in participants inadequately controlled on metformin and pioglitazone. The endpoint description defines what was measured and when. Neither design features nor endpoint labels, however, tell the reader the size or direction of the observed treatment contrast.

The primary endpoint is measured at more than one time point, and the registry does not reduce it to a single comparison. This matters statistically because a result at one visit need not be interchangeable with a result at another. The estimates and their uncertainty need to be interpreted at the specified time points, with the model and analysis plan in view. Those numerical estimates are not reported in the registry record.

The record also highlights how analysis population and missing-data choices shape interpretation. The per-protocol definition excludes randomized participants who did not meet the stated treatment-receipt or protocol criteria, and LOCF was used for imputation. These choices are important to recognize because they define which observations contribute and how missing follow-up values are represented. Without information on missingness and sensitivity analyses, the robustness of the findings cannot be assessed from the registry entries alone.

Multiple endpoints and time points further emphasize the value of a prespecified statistical plan. The record lists 41 outcome measures and 26 analyses, but does not report how multiplicity was handled. This is not evidence that the analysis was inappropriate; it marks a key interpretive detail that is unavailable in the registry information.

Finally, the safety counts show why descriptive event totals and inferential comparisons are not interchangeable. Counts and at-risk denominators convey what was recorded, but do not by themselves quantify uncertainty or establish causation. Taken together, the record is useful for understanding the study’s question and analysis framework while also showing why reported estimates, uncertainty measures, and analytic details are essential to statistical interpretation.

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 registry information does not report numerical primary-outcome estimates for HbA1c change at Week 26 or Week 52. It therefore does not permit assessment of the direction, magnitude, or precision of the primary comparison. The record also states that the outcome includes several categories or time points and is not reduced to one comparison, so no single primary contrast can be assumed from the endpoint description.

The analysis is defined for a per-protocol population, including randomized patients who received at least one dose of double-blind study medication and had no major protocol violations. Results for other analysis populations are not reported in the registry information. This limits comparison of the per-protocol findings with analyses that preserve all randomized participants.

LOCF was used, but the registry does not report the amount, timing, or reasons for missing data, nor whether sensitivity analyses explored different assumptions. The exact statistical model behind the least-squares means is not reported, so its adjustment terms and other modeling choices cannot be assessed. The record also does not describe how multiplicity across 41 outcome measures and 26 analyses was handled.

The safety information gives affected-participant counts and at-risk counts, but the detailed adverse-event definitions, exposure-adjusted summaries, and inferential comparisons are not reported in the information. Group-specific counts for the listed serious-event terms are not reported, and death counts are not reported in the registry record.

Finally, the planning illustration is not a substitute for a trial-specific design justification. The registry information does not report that justification, or numerical primary-outcome results needed to evaluate the study’s achieved precision. These are limitations of the information available in the registry record, not evidence that the trial itself lacked additional analyses or documentation.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Efficacy of Alogliptin and Pioglitazone in Subjects With Type 2 Diabetes Mellitus — NCT00432276.
• Registry record last updated 2013-04-04; status last verified 2013-04.
• Results first posted 2013-04-04.
• 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

The registry describes a completed Phase 3 randomized, double-blind, parallel-group study sponsored by Takeda, with quadruple masking and an enrollment of 803. Its central comparison was alogliptin 25 mg added to pioglitazone 30 mg and metformin versus pioglitazone 45 mg and metformin with alogliptin placebo-matching tablets. The listed regimens could continue for up to 52 weeks.

The primary outcome was change in HbA1c from baseline at Week 26 and Week 52, summarized as a least-squares mean with standard error. The primary population was per-protocol, and LOCF imputation was used. Numerical primary-outcome results and a single summarized comparison are not reported in the registry information.

The record reports serious and other adverse-event counts by group, lists serious-event terms at 0.5% each, and does not report death counts. It lists 41 outcome measures and 26 analyses. The statistical model, multiplicity approach, missing-data patterns, and sensitivity analyses are not reported in the registry record.