Bioequivalence of a FDC Tablet of Linagliptin/Metformin (2.5mg/1000mg) Extended Release in Healthy Subjects.: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Bioequivalence of a FDC Tablet of Linagliptin/Metformin (2.5mg/1000mg) Extended Release in Healthy Subjects. (NCT02084056), a randomized phase 1 study. The primary outcome is Area Under the Concentration-time Curve of Linagliptin in Plasma Over the Time Interval From 0 to 72 Hours (AUC0-72).

2. Clinical Question

The study is framed as a bioequivalence comparison rather than a comparison of clinical outcomes. Its primary endpoint captures drug exposure over a defined period after administration. The record also lists metformin and linagliptin exposure outcomes, but the statistical interpretation below centers on the posted primary linagliptin AUC results. The fed and fasted comparisons are separate registry analyses; their results should not be treated as interchangeable or combined into a single conclusion.

3. Trial Design

This was a randomized, open-label, Phase 1 crossover trial in healthy subjects. The registry lists an enrollment of 74 and marks the study completed. The trial’s stated purpose was to compare a newly developed linagliptin/metformin extended-release fixed-dose combination tablet with the free combination of linagliptin and metformin extended-release tablets.

The arm descriptions set out two treatment sequences for each condition. In the fasted condition, one sequence received the fixed-dose combination followed by the single tablets, while the other received the single tablets followed by the fixed-dose combination. The registry gives corresponding sequences for fed administration. This ordering is consistent with the listed crossover model, in which the treatment comparison is made within a participant across treatment periods. The record does not describe the length of any washout, the period schedule, or how treatment sequence was generated beyond stating that allocation was randomized.

Masking is listed as none, so the trial was open label. The primary endpoint is a pharmacokinetic measure of linagliptin in plasma, rather than a clinical disease outcome. The record describes blood sampling from 1 hour before administration through 72 hours after administration, with multiple scheduled sampling times in between. It reports adjusted geometric means and adjusted intra-individual geometric coefficients of variation for the endpoint.

The posted endpoint summaries have different analysis counts for the treatment conditions: the fasted summaries list 55.0 and 57.0, while the fed summaries list 16.0 and 15.0. These are endpoint analysis counts, not a replacement for the registry’s overall enrollment. The record defines the primary endpoint population using PKS2000 Fast and PKS2000 Fed and an observation and protocol-violation criterion. It does not report why the counts differ between conditions or give a full account of participant flow through the analysis.

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.

Treatment Period 1 (4 Days)

Reasons for not completing this period:

Washout Period 1 (35 Days)

Reasons for not completing this period:

Treatment Period 2 (4 Days)

Analysis population for the primary outcome, as defined in the registry: PKS2000 Fast and PKS2000 Fed, consists of the subjects who provided at least 1 observation for at least 1 primary endpoint without Important protocol violations with respect to the statistical evaluation of pharmacokinetic endpoints.

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: All randomised subjects who received 1 dose of trial medication were included in the treated set.

6. Primary Endpoint

The registry lists the primary outcome as Area Under the Concentration-time Curve of Linagliptin in Plasma Over the Time Interval From 0 to 72 Hours (AUC0-72), measured over 1 hour (h) before drug administration and 20 minutes (min), 40min, 1h,1h 30min, 2h, 3h, 4h, 5h, 6h, 8h, 10h, 12h, 24h,34h, 48h and 72h after drug administration.

The primary endpoint is the area under the plasma concentration-time curve for linagliptin from 0 to 72 hours after administration, reported in nmol*h/L. AUC summarizes exposure across the measured period: it reflects both the concentrations observed and how long they persist during that interval. It is not a direct measure of symptom change or long-term clinical outcome.

Pharmacokinetic AUC is used in bioequivalence work because a formulation that produces similar exposure to a comparator may have similar absorption over the measured interval. That is a general rationale for the endpoint, not a claim that the endpoint alone proves equivalence. Formal bioequivalence interpretation depends on the prespecified margins and analysis approach, and the equivalence margins are not reported in the registry record.

The registry specifies geometric means and geometric coefficients of variation. It further clarifies that the displayed values are adjusted geometric means and adjusted intra-individual geometric coefficients of variation. Geometric summaries are suited to exposure measures that are commonly analyzed on a multiplicative scale. The endpoint was assessed in fasted and fed conditions, with blood sampling listed from 1 hour before administration through 72 hours after administration. The registry also lists AUC from 0 to infinity for linagliptin and metformin as secondary outcomes.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Several statistical ideas help place these pharmacokinetic results in context. The explanations of general statistical concepts in this discussion are general; the specific trial details are identified separately.

What does the crossover design contribute?

In a crossover study, participants receive more than one treatment in a planned sequence. The comparison can therefore use within-participant information, which may reduce the influence of stable differences between participants. That benefit depends on the design and analysis being appropriate, including attention to period and sequence effects and to any carryover between treatments. The registry identifies a randomized two-way crossover model and lists the treatment orders, but it does not report the washout schedule or detailed checks for period, sequence, or carryover effects. Those features should not be assumed from the design label alone.

What does an adjusted geometric mean represent?

General explanation: pharmacokinetic exposure values are often analyzed on a logarithmic scale because values may be positively skewed and treatment effects are often naturally expressed as ratios. A geometric mean is a summary on the original scale after accounting for that log-scale analysis. An adjusted geometric mean is a model-based summary that accounts for terms included in the analysis model. The registry reports adjusted geometric means, but does not describe all model terms or covariates, so the precise adjustment cannot be reconstructed from the registry record.

A ratio compares the estimated exposure under one treatment with the estimated exposure under the other. It is a multiplicative comparison, not a difference in AUC units. The posted ratios are 103.68 for the fasted analysis and 101.63 for the fed analysis. The registry calls these adjusted geometric-means ratios. Its record does not state the ratio’s presentation convention or the formal equivalence margins. The results should therefore be quoted as reported rather than translated into a new scale or interpreted against an unstated threshold.

How should the confidence intervals be read?

General explanation: a confidence interval represents uncertainty around an estimate under a specified statistical model and procedure. A narrower interval indicates greater precision than a wider interval, all else being comparable. A confidence interval is not a probability statement that the true value lies in the particular interval after the data have been observed. For an equivalence analysis, the interval is commonly assessed against prespecified lower and upper bounds. The bounds are essential: an interval’s location and width cannot establish formal equivalence without knowing the acceptance limits.

The registry reports 90% confidence intervals for both primary comparisons. The fasted interval is 100.703 to 106.747; the fed interval is 93.721 to 110.202. The record does not state the margins against which either interval should be judged. Thus, the intervals communicate the reported estimates and uncertainty but do not, on their own, establish whether the study met its formal bioequivalence criterion.

What do the geometric coefficients of variation add?

General explanation: a coefficient of variation describes variability relative to a measure of central tendency. A geometric coefficient of variation is used with log-scale or multiplicative data and provides context for how variable exposure measurements are. The registry describes the posted values as adjusted intra-individual geometric coefficients of variation. It reports 9.1 in both fasted groups and 12.7 in both fed groups. These values are measures of variability, not confidence limits, treatment effects, or proportions of participants with an outcome.

What does the ANOVA label tell the reader?

The registry identifies analysis of variance (ANOVA) for the reported adjusted geometric-means ratios. In general, ANOVA is a modeling framework for comparing group means; pharmacokinetic analyses may use transformed measurements and a design-specific model. The method label alone does not reveal the model terms, transformation details, handling of incomplete observations, or precise hypothesis test. Those details are not reported in the registry record. The reported p values are p < 0.001 for the fasted comparison and < 0.001 for the fed comparison, but the record does not define the null hypothesis attached to them. They should not be given a more specific interpretation than the registry supports.

Why are endpoint analysis counts relevant?

The primary endpoint summaries list counts by treatment and condition, and these counts vary across the posted comparisons. In a crossover analysis, the usable information may depend on which participants have endpoint observations and satisfy analysis criteria. The registry describes a pharmacokinetic analysis population requiring at least 1 observation for at least 1 primary endpoint and no important protocol violations relevant to pharmacokinetic statistical evaluation. It does not give a detailed account of missing observations, exclusions, or the reason for the different counts. The overall enrollment should therefore not be treated as the analysis count for every endpoint comparison.

Finally, statistical significance and equivalence answer different questions. A p value assesses a specified hypothesis under a statistical model. Equivalence asks whether the treatment difference is sufficiently constrained within prespecified limits. General statistical interpretation requires both the estimate and its interval, as well as the criterion fixed for the study. The registry record contains estimates, intervals, and p values, but does not report the margins or enough detail to reconstruct the full decision process.

10. Analyses Reported by the Sponsor

The registry reports sponsor analyses using ANOVA and adjusted geometric-means ratios for the primary linagliptin AUC comparisons. The fasted estimate is 103.68, with a 90% confidence interval from 100.703 to 106.747; the fed estimate is 101.63, with a 90% confidence interval from 93.721 to 110.202. The record also reports p < 0.001 and p < 0.001, respectively.

No separate recalculated estimate is reported in the registry record for these comparisons. The posted summary statistics are not accompanied by enough detail to reproduce the sponsor’s adjusted model-based analyses independently. The registry does not flag a discrepancy between sponsor and recalculated results, but that absence is not an independent validation of the analysis. The sponsor’s reported analysis should be read as the available analysis, with its interpretation limited by the missing equivalence margins and incomplete model details.

The reported p values do not replace the equivalence assessment. A reader cannot determine from the record whether the confidence intervals met the study’s prespecified acceptance limits, because those limits are not reported. The sponsor-reported estimates are therefore useful for describing the observed exposure ratios and their uncertainty, but the registry record does not support reconstructing the full formal bioequivalence decision.

11. Multiplicity, Missing Data and Analysis Populations

The registry lists 6 outcome measures and 12 analyses, including the primary linagliptin AUC outcome and secondary outcomes for linagliptin AUC from 0 to infinity and metformin AUC from 0 to infinity. The record does not report a multiplicity adjustment, a hierarchy for testing outcomes, or a rule for interpreting results across the analyses. It is therefore unclear whether the reported p values were intended as confirmatory tests within a prespecified family or as descriptive components of a broader pharmacokinetic assessment. The p values should not be treated as if the record documented a particular multiplicity strategy.

The registry defines the primary endpoint population as PKS2000 Fast and PKS2000 Fed: subjects who provided at least 1 observation for at least 1 primary endpoint and had no important protocol violations with respect to statistical evaluation of pharmacokinetic endpoints. This gives a partial description of the analysis population, but it does not specify all exclusions, the number excluded for each reason, or the exact participant flow into each analysis. The posted endpoint counts differ by condition and treatment summary. The reason for that variation is not reported in the registry record.

Missing-data handling is also not reported in the registry record. The record does not explain whether observations were omitted, imputed, or handled through another model-based approach, nor does it describe sensitivity analyses. Because the endpoint is based on serial concentration measurements, incomplete sampling could affect the estimable exposure summary; the extent of any such issue cannot be determined from the record.

The registry does not report interim analyses or a sample-size justification. It is not possible to infer the study’s planning assumptions or whether any analysis decisions were made before or after reviewing the data. These omissions limit assessment of the analysis plan’s completeness, but they do not establish that a particular procedure was or was not performed.

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 registry lists serious adverse events as affecting 0 participants in each of the four treatment-condition groups. The corresponding numbers at risk are 56 for the fasted single-tablet combination, 58 for the fasted fixed-dose combination, 16 for the fed single-tablet combination, and 16 for the fed fixed-dose combination. These are the registry’s adverse-event group counts and should be kept distinct from the pharmacokinetic endpoint analysis counts.

For other adverse events, the registry reports 20 affected participants among 56 at risk in the fasted single-tablet group, 24 among 58 in the fasted fixed-dose group, 8 among 16 in the fed single-tablet group, and 5 among 16 in the fed fixed-dose group. These counts do not establish whether an event was caused by treatment, and the registry record does not provide event-level descriptions in the registry record. The serious-event term list is empty.

Deaths are not reported in the registry record. The absence of reported serious adverse events in these groups should be read as a count in this trial record, not as proof that a treatment has no serious risks. This is a Phase 1 study in healthy subjects, and the available counts do not provide a complete basis for comparing safety or drawing conclusions beyond the recorded events.

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

The primary results describe plasma linagliptin exposure after a fixed-dose combination tablet and after the corresponding single-tablet combination. The registry reports adjusted geometric means, adjusted geometric-mean ratios, confidence intervals, and adjusted intra-individual geometric coefficients of variation separately for fasted and fed conditions. Those data characterize the estimated exposure comparison and its statistical precision in the analyzed subjects.

They do not, by themselves, establish a clinical benefit or a treatment effect on health outcomes. The primary endpoint is a pharmacokinetic measure, not a symptom, complication, or long-term outcome. Nor do the reported numbers alone establish formal bioequivalence. To make that determination, the estimate and interval must be interpreted against prespecified equivalence margins, which are not reported in the registry record. The model specification and the precise meaning of the reported p values are also not fully described.

The fasted and fed analyses should not be collapsed into a single result. Their endpoint counts and reported intervals differ, and the record provides no combined estimate or direct statistical comparison of food conditions. The results also apply to the population and analysis criteria described in the registry, not automatically to other populations or settings.

The record lists secondary exposure outcomes for linagliptin and metformin, but their numerical results are not available in the registry record used. No conclusion about those outcomes should be inferred from the primary linagliptin result. Similarly, the adverse-event counts are descriptive registry entries and do not establish causality or characterize all possible harms.

The most defensible reading is therefore narrow: the registry reports close adjusted geometric-mean ratios for the stated comparisons, with confidence intervals of differing precision. The information is useful for understanding the pharmacokinetic comparison, while important elements required for a complete formal assessment—including equivalence limits and full analytic details—are not reported in the registry record.

14. Why This Trial Matters Statistically

This trial illustrates why bioequivalence results require more than a point estimate. The ratio summarizes the relative exposure estimated for the two formulations, while the confidence interval shows uncertainty around that estimate. For a formal equivalence question, the decision depends on whether the interval satisfies prespecified bounds. Without those bounds, even a ratio that appears close to the comparator cannot be translated into a definitive equivalence conclusion.

The crossover structure is also statistically relevant. General explanation: comparing treatments within the same participant can account for stable participant-level variation, which is useful when pharmacokinetic measurements vary among people. At the same time, the interpretation depends on appropriate conduct and analysis across periods. The registry identifies randomized treatment sequences and a crossover model, but does not provide all design details needed to assess period, sequence, or carryover considerations.

Separate fasted and fed results make the condition-specific evidence visible rather than obscuring it in a single pooled summary. The endpoint counts and confidence intervals differ between conditions, so the uncertainty should be considered separately. The record does not report a direct test of whether food changes the relative comparison, and no such conclusion should be inferred from a visual or numerical contrast between the estimates.

Finally, this record demonstrates the difference between a stated study purpose and what can be verified from posted results. The purpose is to demonstrate bioequivalence, but the available registry entries do not include the equivalence margins or a complete description of the analysis model. Statistical interpretation depends on those details, not only on the wording of the study objective. The posted results remain informative as exposure estimates, while the formal decision cannot be fully audited from the registry record alone.

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 record does not report the prespecified equivalence margins. This is a central limitation for interpreting a trial whose stated purpose is bioequivalence: the reported confidence intervals cannot be assessed against an acceptance criterion that is not available. The p values are posted, but the record does not specify their null hypothesis, so their role in the formal decision is unclear.

The analysis description is incomplete. ANOVA is named, and the results are described as adjusted geometric means and adjusted intra-individual geometric coefficients of variation, but the model terms and adjustment details are not reported. The record also does not describe the washout schedule, period structure, or assessment of sequence and carryover effects. These missing details prevent a full independent evaluation of the crossover analysis.

The endpoint analysis counts vary between conditions and treatment summaries. The registry describes eligibility for the PKS2000 Fast and PKS2000 Fed populations, but does not explain the different counts, provide a detailed flow of participants, or report how missing pharmacokinetic observations were handled. The overall enrollment of 74 should not be assumed to represent the number contributing to every endpoint analysis.

The registry lists multiple outcomes and analyses but does not report a multiplicity strategy or testing hierarchy. It also does not report interim analyses or sample-size justification. Numerical results for the secondary outcomes are not included in the registry record for this interpretation. These limitations constrain conclusions to the posted primary exposure summaries and the adverse-event counts, without filling gaps through assumptions about the protocol or analysis plan.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Bioequivalence of a FDC Tablet of Linagliptin/Metformin (2.5mg/1000mg) Extended Release in Healthy Subjects. — NCT02084056.
• Registry record last updated 2016-08-04; status last verified 2016-06.
• Results first posted 2016-08-04.
• 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.

19. Record Summary

The registry record describes a completed, randomized, open-label, Phase 1, two-way crossover trial in healthy subjects, with enrollment of 74. It compares a linagliptin/metformin extended-release fixed-dose tablet with the same drugs given as single tablets in fasted and fed conditions. The primary endpoint is linagliptin plasma AUC from 0 to 72 hours.

For the fasted comparison, the reported adjusted geometric-means ratio is 103.68, with a 90% confidence interval from 100.703 to 106.747. For the fed comparison, the ratio is 101.63, with a 90% confidence interval from 93.721 to 110.202. The registry also reports p < 0.001 and p < 0.001, respectively.

The record does not report the equivalence margins, full model specification, or missing-data handling. The posted findings therefore describe exposure estimates and uncertainty but do not permit a complete independent assessment of the formal bioequivalence decision.