Two-way Crossover Study in Healthy Male and Female Subjects to Evaluate the Bioequivalence of Jentadueto®.: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Two-way Crossover Study in Healthy Male and Female Subjects to Evaluate the Bioequivalence of Jentadueto®. (NCT01947153), a randomized phase 1 study. The primary outcome is Linagliptin: AUC 0-72 (Area Under the Concentration-time Curve of the Analyte in Plasma Over the Time Interval From 0 to 72 Hours).

2. Clinical Question

This is a pharmacokinetic comparison in healthy participants, rather than a study of clinical outcomes or treatment effects in people with a disease. The primary endpoint is a measure of drug exposure over a stated interval. The registry also lists outcomes for other linagliptin and metformin exposure measures, but values for some comparisons are missing or not estimable in the posted results. The findings therefore address the recorded formulation comparison and endpoint, not broader questions about clinical benefit or safety in other populations.

3. Trial Design

The registry describes a completed Phase 1, randomized, unmasked, two-way crossover study sponsored by Boehringer Ingelheim. The listed enrollment is 36, and the study population is healthy male and female participants. The record gives a start date of 2013-11 and a primary completion date of 2014-01.

Participants were compared across two formulations: a fixed-dose combination of linagliptin and metformin and a free combination of linagliptin and metformin tablets. The official study title specifies the respective doses as two fixed-dose tablets containing linagliptin 2.5 mg and metformin 500 mg each, compared with linagliptin 5 mg and metformin 1000 mg. The comparison was conducted under fasting conditions.

A crossover design means participants receive more than one study condition in a sequence, allowing comparison of measurements from the same participant under different conditions. This can be useful in pharmacokinetic studies because within-person comparisons may account for some differences between participants. The registry labels the model as crossover, but details of treatment sequences, sequence allocation, washout procedures, and period-specific analysis are not reported in the registry record. The masking field is listed as none.

The primary outcome was linagliptin AUC from 0 to 72 hours. The registry states that blood sampling was planned from 2 hours before drug administration through 72 hours after administration, with multiple measurements at specified times. The primary endpoint analysis population was a linagliptin pharmacokinetic set with eligibility based on having at least one relevant observation and no important protocol violations for the statistical evaluation. Further details about exclusions and the number analyzed for the primary endpoint are not reported in the registry record. The trial is marked completed.

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.

Test Period 1

Test Period 2

Analysis population for the primary outcome, as defined in the registry: subject set for the evaluation of pharmacokinetic endpoints of linagliptin (PKS-L): includes all subjects of the TS who provided at least 1 observation for at least 1 primary endpoint for linagliptin and had no important protocol violations with respect to the statistical evaluation of pharmacokinetic endpoints of linagliptin

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: Treated set (TS): includes all subjects who had been dispensed study medication and were documented to have taken at least 1 dose of study medication.

6. Primary Endpoint

The registry lists the primary outcome as Linagliptin: AUC 0-72 (Area Under the Concentration-time Curve of the Analyte in Plasma Over the Time Interval From 0 to 72 Hours), measured over 2 hours (h) before drug administration and 20 minutes (min), 40min, 1h, 1h 30min, 2h, 3h, 4h, 6h, 8h, 10h, 12h, 24h, 34h, 48h and 72h after drug administration.

The primary endpoint is linagliptin AUC from 0 to 72 hours, measured in nmol*h/L. AUC, or area under the concentration-time curve, represents drug concentration accumulated over a specified interval. In this case, concentrations are followed from administration through 72 hours. The registry describes sampling from 2 hours before administration and at multiple times after administration, ending at 72 hours.

Pharmacokinetic exposure measures such as AUC are used to compare how much of a drug reaches and remains in the bloodstream over time under different formulations. For a bioequivalence question, comparing exposure is relevant because different formulations can alter absorption and the resulting concentration profile. The registry reports geometric means for each formulation and an ANOVA-based geometric mean ratio with a 90% confidence interval.

The endpoint is not a direct measure of symptom improvement, disease control, or a clinical event. It describes plasma exposure to linagliptin in the study’s healthy-participant population. The posted primary endpoint population is the linagliptin pharmacokinetic set, which excludes subjects with important protocol violations relevant to the endpoint evaluation. The number included in that set is not reported in the registry record beyond the arm-level counts in the posted results.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does the geometric mean ratio describe?

The geometric mean ratio compares the estimated geometric mean exposure for one formulation with that for the other. Geometric means are commonly used for pharmacokinetic measures because exposure values can be positively skewed: a small number of relatively high values can pull an ordinary arithmetic mean upward. A geometric-scale analysis works with logarithms of the measurements and then expresses the comparison back on the original ratio scale.

For this trial, the registry reports an ANOVA analysis with a geometric mean ratio of 102.92. It also reports a 90% confidence interval from 98.370 to 107.688. The ratio is a relative comparison, whereas the arm-level geometric means of 293.0 and 288.0 nmol*h/L describe exposure in each formulation. The confidence interval quantifies uncertainty in the estimated ratio under the analysis model. It is not a range containing a stated proportion of individual participants’ values, nor does it show the full distribution of exposure.

Why does the interval need a decision margin?

Equivalence is a margin-based question. Before analysis, a study generally defines acceptable bounds for the ratio; the result is then assessed by asking whether the confidence interval falls entirely within those bounds. A confidence interval can be estimated without the reader knowing the decision limits, but it cannot by itself show whether a prespecified equivalence rule was satisfied. The registry record does not report the limits, so no conclusion about their relationship to the posted interval can be drawn from the record.

The reported p-value, p < 0.001, is also part of the posted analysis. A p-value depends on a specific null hypothesis and test. The registry labels the analysis as non-inferiority or equivalence but does not explain which hypothesis the reported p-value addresses. It should not be treated as a stand-alone probability that the formulations are equivalent, or as the probability that the observed result arose by chance. The interval and the decision criterion are central to understanding an equivalence analysis.

What does ANOVA contribute?

ANOVA, or analysis of variance, is a framework for comparing outcomes across conditions while accounting for modeled sources of variation. In a crossover study, a suitable analysis can account for the repeated measurements on participants and for design features such as treatment and period. The registry names ANOVA but does not report the full model specification, its included terms, assumptions, or diagnostics. Details of how sequence, period, participant, or carryover effects were handled are not reported in the registry record and should not be inferred from the method label alone.

How should the geometric coefficients of variation be read?

The arm-level geometric coefficients of variation are 23.4 for the fixed-dose combination and 22.9 for the free combination. A coefficient of variation describes dispersion relative to the scale of the measurements. On a geometric scale, it conveys variability in a way that aligns with a log-scale analysis. It is not a confidence interval, and it does not identify the range of participant values. Greater variability can make a comparison less precise because the estimate has more uncertainty, although the actual precision also depends on the analysis design and available observations.

Pharmacokinetic measurements can show multiplicative variability: an individual’s exposure may differ proportionally rather than by a constant absolute amount. Log-scale methods can be useful in that setting. The registry does not provide raw observations, distributions, diagnostic plots, or a detailed account of transformations, so those features cannot be independently assessed from the posted summary statistics.

Why does crossover structure matter?

In a crossover design, each participant can contribute measurements under more than one condition. Comparing a participant with themself may reduce the influence of stable differences between people. At the same time, the analysis must respect the fact that observations from the same person are related rather than independent. Treatment order, study period, and any residual effect from an earlier treatment can also matter. The registry identifies a two-way crossover, but does not report sequence details, washout information, or the full statistical model. The summary should therefore be understood as the posted analysis, not as an independently verified reconstruction of the design-specific model.

The posted arm-level sample sizes are 36 and 34. They do not by themselves describe the number of independent participants contributing paired observations to the comparison. The analysis population definition indicates that subjects needed relevant pharmacokinetic observations and no important protocol violations for the statistical evaluation. The registry does not give a participant flow or a detailed explanation for the difference in counts.

What can planning calculations say?

Planning illustrations based on an observed effect are not observed power. They show how a hypothetical future study might be sized under stated assumptions, such as a two-sided significance level of 0.05, equal allocation, and a normal approximation. The illustrations are conditional on the chosen effect fraction and assumptions; they do not establish the power of this completed trial or explain how its original sample size was selected. The registry record does not report the trial’s own sample-size justification or design assumptions.

10. Analyses Reported by the Sponsor

The sponsor’s posted comparison uses ANOVA and reports a geometric mean ratio of 102.92 with a 90% confidence interval from 98.370 to 107.688 and p < 0.001. The arm-level geometric means are 293.0 nmol*h/L for the fixed-dose combination and 288.0 nmol*h/L for the free combination. These are the registry’s reported results.

The registry record does not provide a separate recalculated estimate for this endpoint or a discrepancy between the sponsor analysis and a recalculation. Consequently, there is no numerical recalculation to compare with the sponsor’s result. The sponsor’s own analysis is the relevant reported inferential analysis. Its ratio and interval should be interpreted in light of the study’s equivalence or non-inferiority criterion; the decision limits and fuller model details are not reported in the registry record.

The arm-level summaries are descriptive and do not reproduce a crossover ANOVA. A calculation based only on summaries would not necessarily account for the within-participant structure or the analysis model. No separate recalculation should be treated as replacing the sponsor’s analysis. The result supports description of the posted exposure comparison, but the record does not give enough detail to independently verify every modeling choice or the equivalence decision rule.

11. Multiplicity, Missing Data and Analysis Populations

The registry lists seven outcome measures and seven analyses. It identifies the linagliptin AUC from 0 to 72 hours endpoint as primary and lists three secondary outcomes, including AUC measures for linagliptin and metformin. Some comparisons are marked as having values missing or not estimable. The registry does not provide a complete set of numerical results for every listed outcome.

When a study evaluates multiple outcomes, the chance of at least one apparently notable result can increase if each is tested separately. A prespecified hierarchy, adjustment, or other multiplicity strategy can help define how results should be interpreted. The registry record does not report whether a multiplicity adjustment was planned or used, how outcomes were prioritized beyond identifying the primary endpoint, or how the posted p-value relates to any broader testing plan. No adjustment should be assumed.

The primary endpoint population is described as the linagliptin pharmacokinetic set. It includes trial subjects with at least one observation for at least one primary linagliptin endpoint and no important protocol violations relevant to the statistical evaluation. This gives some information about eligibility for analysis, but the registry record does not report a full account of exclusions, missing-data handling, imputation, or sensitivity analyses. The specific reasons that arm-level endpoint counts are 36 and 34 are also not reported.

Missing or not-estimable outcome values limit the scope of interpretation. They do not reveal whether data were unavailable, excluded under analysis rules, or otherwise could not be estimated; those details are not reported in the registry record. Similarly, the registry does not report interim analyses or a detailed statistical analysis plan. Conclusions should remain tied to the available endpoint-specific results rather than extending to unreported outcomes or unreported analytic procedures.

12. 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 non-serious adverse event terms

The posted safety figures list 36 participants at risk in each formulation group. Serious adverse events affected 0 in the fixed-dose combination group and 0 in the free-combination group. No top serious adverse-event terms are listed. Death counts are not reported in the registry record.

For other adverse events, the registry lists 1 affected participant among 36 at risk for the fixed-dose combination and 4 affected participants among 36 at risk for the free combination. These counts describe what is posted; they do not establish whether an event was caused by a study drug, and the record does not provide event-level details or severity information for these other events.

These figures should not be used to infer that the formulations have equivalent safety. The study enrolled a limited healthy-participant population and its primary focus was pharmacokinetic bioequivalence. Zero recorded serious adverse events in the listed groups is a description of this study’s posted observations, not evidence that serious events cannot occur. The registry record does not report additional safety context needed to assess differences or risks beyond these counts.

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

The main result is a comparison of linagliptin exposure between the fixed-dose combination and the free combination under fasting conditions in healthy male and female participants. For the primary AUC from 0 to 72 hours endpoint, the posted analysis reports a geometric mean ratio of 102.92 and a 90% confidence interval from 98.370 to 107.688. The study is labeled as a bioequivalence study, and the analysis is labeled as non-inferiority or equivalence.

The reported interval gives information about the precision of the estimated ratio under the analysis model, but a bioequivalence decision depends on the limits specified for the study. Because those limits are not reported in the registry record, the interval cannot be definitively classified against the study’s decision rule using this record alone. The reported p-value, p < 0.001, does not replace that missing information, and its precise hypothesis is not described.

The result does not show whether the formulations produce the same clinical outcomes, nor does it address treatment effectiveness in people with a disease. It concerns a pharmacokinetic measure of linagliptin exposure in a healthy-participant setting. The intervention includes metformin as well, but the primary endpoint is linagliptin AUC. Other listed pharmacokinetic measures are not all accompanied by estimable posted comparisons.

The findings also do not establish that there are no safety differences. The registry lists no serious adverse events in either group and lists other adverse events in both, but these observations are limited to this study and the record’s available detail. They cannot establish what would occur in a different population or over a different period.

In short, the registry presents a sponsor-reported pharmacokinetic comparison with a confidence interval and p-value. The absence of equivalence margins, full model details, and complete outcome results limits how far the result can be interpreted. It is appropriate to describe the reported estimate and its uncertainty, while distinguishing that description from an independently verified conclusion about bioequivalence or clinical effects.

14. Why This Trial Matters Statistically

This study illustrates why a pharmacokinetic comparison cannot be reduced to comparing two arm-level means. The primary endpoint is an exposure measure, and the central inferential result is a geometric mean ratio with a confidence interval. That ratio-based estimate aligns with a question about relative exposure, while its interval communicates uncertainty that is not captured by the point estimate alone.

The crossover structure is also statistically important. When participants contribute measurements under multiple conditions, within-person comparisons can be informative, but the repeated measurements need to be modeled appropriately. The registry names an ANOVA analysis and a two-way crossover design, while leaving details of the model, periods, and sequences unreported. This distinction is a useful reminder that a method label is not a full analysis specification.

The record also demonstrates the role of decision criteria in equivalence studies. An estimate and confidence interval need to be assessed against prespecified limits; without those limits, the result can be described precisely but the decision rule cannot be reconstructed from the registry. A small p-value should not be substituted for that criterion when the hypothesis being tested is not stated.

Finally, the endpoint counts, incomplete outcome results, and limited safety information show why statistical interpretation depends on the completeness of reporting as well as on the numerical estimate. The reported endpoint has a clear pharmacokinetic meaning, but it cannot answer broader clinical questions or fill gaps in unreported analysis details. Keeping those boundaries explicit makes the result more useful and less likely to be overinterpreted.

15. 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 identifies the study as randomized and crossover, but does not report the treatment sequences, allocation details, washout interval, period structure, or full ANOVA specification. These omissions limit independent assessment of how the design was implemented and how the analysis accounted for repeated measurements and possible period effects.

The record does not report the equivalence margins used for the primary comparison. Although the sponsor analysis reports a geometric mean ratio and a 90% confidence interval, the decision criterion cannot be checked against the posted interval without those bounds. The precise hypothesis associated with p < 0.001 is also not described.

The primary endpoint counts are 36 and 34, and the record does not explain why they differ. The endpoint population has eligibility requirements relating to pharmacokinetic observations and protocol violations, but full participant disposition, exclusion reasons, and missing-data methods are not reported. The record also does not provide participant-level data or sufficient detail to reproduce the crossover analysis.

Not all listed outcomes have estimable posted values, and the registry does not describe multiplicity control, interim analyses, or the complete statistical analysis plan. Safety reporting is limited to the posted event counts; event-level details and death counts are not reported. The study’s healthy-participant setting and pharmacokinetic focus also constrain what can be inferred about clinical outcomes or other populations. These limitations do not negate the reported estimate, but they narrow the conclusions that can be drawn from the registry record.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Two-way Crossover Study in Healthy Male and Female Subjects to Evaluate the Bioequivalence of Jentadueto®. — NCT01947153.
• Registry record last updated 2015-02-25; status last verified 2015-02.
• Results first posted 2015-02-25.
• 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.

19. Record Summary

NCT01947153 is a completed Phase 1, randomized, unmasked, two-way crossover study sponsored by Boehringer Ingelheim. It enrolled 36 healthy participants and compared a fixed-dose combination of linagliptin and metformin with the free combination under fasting conditions.

The primary endpoint was linagliptin AUC from 0 to 72 hours. The posted ANOVA analysis reports a geometric mean ratio of 102.92, a 90% confidence interval from 98.370 to 107.688, and p < 0.001. Equivalence margins and full model details are not reported in the registry record. The registry also lists other pharmacokinetic outcomes, some of which are missing or not estimable, and gives limited adverse-event counts. Interpretation is therefore best confined to the reported primary exposure comparison and the information explicitly documented in the record.