The Effect of Gemfibrozil, Ketoconazole and Clarithromycin on the Amount of LY2409021 in the Bloodstream: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for The Effect of Gemfibrozil, Ketoconazole and Clarithromycin on the Amount of LY2409021 in the Bloodstream (NCT01836198), a randomized phase 1 study. The primary outcome is Pharmacokinetics: Area Under the Concentration Curve From Zero to Infinity (AUC[0-∞]) of LY2409021.

2. Clinical Question

In pharmacokinetic studies, a probe drug can be used to examine whether another drug’s concentration-time profile changes during coadministration. The stated purpose was to examine how much LY2409021 is found in the bloodstream and how long the body takes to get rid of it. AUC(0-∞) addresses overall exposure across the concentration-time profile; it is not a direct measure of a clinical outcome.

The registry identifies healthy volunteers as the condition and describes a Phase 1 study. It does not report a clinical efficacy question, a disease population, or a clinical outcome. The primary endpoint population is specified as participants with at least 1 dose and evaluable AUC(0-∞) data, rather than every enrolled participant regardless of endpoint availability.

3. Trial Design

The registry describes NCT01836198 as a completed, randomized, unmasked, parallel Phase 1 study sponsored by Eli Lilly and Company. It lists enrollment of 30, a start date of 2013-05, and primary completion in 2013-11. The participants were healthy volunteers. Although the study is classified as parallel, the study description also includes repeated periods and observations for some participants; the registry’s arm labels should therefore be read in the context of those parts and cohorts.

Part A had two cohorts. In each cohort, Period 1 involved a single 20-mg oral dose of LY2409021 on Day 1. In Cohort 1, Period 2 involved 600 mg gemfibrozil orally in the morning and evening on Days 1-20, followed by a morning dose on Day 21; a single 20-mg oral dose of LY2409021 was given on Day 4. In Cohort 2, Period 2 involved 400 mg ketoconazole orally once daily on Days 1-21, with a single 20-mg oral dose of LY2409021 on Day 4.

Part B was limited to participants who successfully completed Part A. It involved 500 mg clarithromycin orally twice daily on Days 1-21 and a single 20-mg oral dose of LY2409021 on Day 4. The registry describes screening within 27 days of the start of the study for each part, and gives durations of 51 days for Part A and 29 days for Part B.

The posted arm counts represent observations in these study conditions, including data from participants who contributed to different periods or parts. They should not be added as though every arm necessarily represents a separate set of unique participants. The registry record does not report further details of the randomization sequence or allocation concealment.

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: This study was conducted in 2 parts at a single center as an inpatient and outpatient study. Part A was an open-label, fixed sequence, 2-period, 2-cohort study. Part B was an open-label, 1-period study. Participants who completed Part A were eligible to participate in Part B.

Part A, Period 1

Part A, Period 2

Reasons for not completing this period:

Part B

Analysis population for the primary outcome, as defined in the registry: All participants who received at least 1 dose of study drug and had evaluable AUC(0-∞) data.

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: All participants enrolled in the study.

6. Primary Endpoint

The registry lists the primary outcome as Pharmacokinetics: Area Under the Concentration Curve From Zero to Infinity (AUC[0-∞]) of LY2409021, measured over Predose and 0.5, 1, 2, 4, 8, 12, 24, 48, 72, 96, 144, 216, and 336 hours postdose (and 408, 504, and 576 hours postdose in Part A, Period 2 and Part B only).

The primary endpoint is the area under the concentration curve from zero to infinity, written AUC(0-∞), for LY2409021. Its unit is nanogram*hours per milliliter (ng*h/mL). The registry describes the endpoint as a geometric mean and reports a geometric coefficient of variation as its dispersion measure.

In general, an area under a concentration-time curve represents overall drug exposure over the measured time course. Extending the area to infinity means that the endpoint is intended to include exposure beyond the last measured concentration, using an extrapolated tail. The registry lists predose and postdose sampling times through 336 hours, and additional times through 576 hours in Part A, Period 2 and Part B. It does not describe the specific calculation method used for the extrapolated portion.

AUC is commonly used in pharmacokinetic research because it summarizes the concentration profile across time rather than relying on a single measurement. In general, comparing AUC under different coadministration conditions can help characterize whether exposure changes. That is a pharmacokinetic comparison, not by itself evidence about a clinical benefit or harm. The registry does not state why this endpoint was selected over other pharmacokinetic measures, and it does not post secondary outcome results.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary results use geometric means and geometric coefficients of variation. The registry record does not post an analysis description for the primary endpoint, so the statistical procedures actually used for estimating or comparing the arm results are not reported in the registry record. The concepts below explain how to read the posted summaries without attributing an undocumented analysis method to the investigators.

What does AUC measure?

In general, pharmacokinetic sampling records drug concentrations at specified times after dosing. A concentration-time curve provides a way to describe how concentrations change over time. The area under that curve summarizes exposure across the interval, rather than describing concentration at only one time point. For this study, the named endpoint is AUC(0-∞), expressed in ng*h/mL.

The registry lists predose measurements and postdose sampling at 0.5, 1, 2, 4, 8, 12, 24, 48, 72, 96, 144, 216, and 336 hours. It also lists 408, 504, and 576 hours in Part A, Period 2 and Part B only. In general, measurements over time contribute to the observed portion of an AUC, while the “to infinity” portion requires extrapolation beyond the final measurement. The record does not report how that extrapolation was performed or what criteria were used to judge an AUC estimate evaluable.

Why report a geometric mean?

In general, pharmacokinetic measurements are often described on a multiplicative scale. A geometric mean is calculated by taking the mean after logarithmic transformation and then transforming back to the original scale. It summarizes the center of a distribution while respecting multiplicative rather than additive variation. The geometric means posted for the arms are expressed in the endpoint’s original unit, ng*h/mL.

Geometric means do not describe every participant’s exposure, and a group mean is not the same as a participant-level result. A difference between two posted geometric means is also not automatically a tested treatment effect. The registry reports arm-level summaries but no contrast estimate, standard error, confidence interval, or p-value for the comparisons of interest. No recalculated comparison is posted in this record.

What does the geometric coefficient of variation convey?

In general, a coefficient of variation expresses spread relative to a measure of central tendency. The geometric coefficient of variation applies this idea to a distribution analyzed on a logarithmic scale. Larger values indicate more relative variability in the observations; they do not show the precision of an estimated between-condition contrast. The values posted for the study are 31.0, 35.0, 21.0, 23.0, 25.0, and 28.0 for the respective arms. The registry does not state whether any transformation or other adjustment was used for the endpoint analysis.

What can an interval estimate tell a reader?

In general, a confidence interval gives a range of values compatible with a statistical estimate under the method’s assumptions. Its width reflects uncertainty as well as the scale of the estimate. It can help distinguish a relatively precise estimate from one with substantial uncertainty. No confidence interval is posted for the arm means or for comparisons between conditions, so the precision of a contrast cannot be assessed from the registry summaries alone. A geometric coefficient of variation should not be read as a substitute for such an interval.

How should sample-size illustrations be read?

The planning illustrations described for this record use the observed effect, a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation; the effect to detect is a stated fraction of the observed effect. These assumptions describe illustrations for planning, not the trial’s original design. The registry record does not report the trial’s own sample-size justification. A sample-size illustration based on an observed effect is not observed power and does not show that the study had a particular probability of detecting an effect.

For this study, the posted information does not include the numerical planning results or a single comparison to which every arm belongs. Consequently, the design assumptions alone do not establish the required sample size for a future study or the precision achieved in this one. Such conclusions would require the relevant comparison, its variability, and a prespecified analysis approach.

10. Multiplicity, Missing Data and Analysis Populations

The registry record does not report whether the analyses addressed multiplicity. This study includes three probe-drug conditions and multiple pharmacokinetic sampling times, but the primary endpoint is AUC(0-∞); the record does not state a hierarchy of comparisons, an adjustment method, or a strategy for controlling the chance of false-positive findings across analyses. Because no p-values or contrast tests are posted, the role of multiplicity in interpreting statistical significance cannot be assessed.

Missing-data handling is not reported in the registry record. The primary population is defined as participants who received at least 1 dose of study drug and had evaluable AUC(0-∞) data. This definition identifies the participants contributing to the endpoint, but does not explain how many participants were excluded for unavailable or nonevaluable measurements, why data may have been unavailable, or how incomplete concentration profiles were handled. The registry does not describe imputation or sensitivity analyses.

The randomized designation is reported, but the analysis population beyond the endpoint-specific definition is not. The record does not say whether analyses followed an intention-to-treat principle, used a per-protocol population, or applied another population rule. It also does not report the statistical model, covariates, handling of repeated observations, or adjustments for cohort or period.

There are no posted secondary outcome results and no inferential analyses reported. The record therefore supports description of the posted endpoint summaries, but not an assessment of a prespecified multiplicity plan, missing-data robustness, or the consistency of alternative analysis populations.

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 non-serious adverse event terms

The registry lists zero participants with serious adverse events in each displayed group. The serious-event denominators are 15 in each Part A group and 12 in each Part B group. These are the counts recorded for the listed exposures; they do not establish that an event could not occur in a larger or different population. Death counts are not reported in the registry record.

The registry also lists other, non-serious adverse-event counts separately from serious events. For example, the LY2409021-only group in Part A, Cohort 1 has 6 participants with other adverse events, while the LY2409021-plus-ketoconazole group has 9. The distinction matters: an “other adverse event” count is not a serious adverse event count, and these counts should not be combined or relabeled as serious events.

These safety figures are counts of participants affected within the listed groups, not counts of individual event episodes. The record does not provide serious adverse-event terms, and the top serious terms list is empty. It also does not describe event severity beyond the serious/non-serious grouping in these summaries, timing, relatedness assessment, or detailed follow-up. With a small healthy-volunteer study, the observed counts are descriptive and should not be treated as a comprehensive characterization of uncommon safety outcomes.

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

The results establish that the registry posts pharmacokinetic summaries for LY2409021 alone and with gemfibrozil, ketoconazole, or clarithromycin in the study’s specified parts and cohorts. The geometric means differ across the listed conditions, with the ketoconazole combination having the largest posted geometric mean among these arms. The geometric coefficients of variation describe relative variability in the observations. These are descriptive findings from the posted results.

The record does not provide statistical tests, confidence intervals, or adjusted comparisons. It therefore does not show how precisely the between-condition differences were estimated or whether a formal statistical criterion was met. In particular, a larger observed geometric mean is not, by itself, proof of a statistically reliable difference. Comparisons also need to respect the study structure: the gemfibrozil and ketoconazole arms are in separate Part A cohorts, and the Part B LY2409021-only results are data from Part A participants.

The endpoint is pharmacokinetic exposure, not a measure of clinical efficacy, patient experience, or a clinical outcome. Participants were healthy volunteers, and the registry does not report results from a disease population. The findings therefore describe the study’s drug-concentration summaries; they do not establish what would happen in patients or answer a treatment decision.

The record also does not report the protocol’s statistical analysis plan, missing-data procedures, multiplicity approach, or sample-size justification. Without those details and inferential estimates, the summaries cannot establish the robustness of a comparison or the level of uncertainty around it. The results should be read as a limited description of observed AUC values under the listed conditions, not as a definitive ranking of clinical effects.

13. Why This Trial Matters Statistically

This study illustrates a common statistical feature of early pharmacokinetic research: the central question concerns changes in drug exposure, rather than a clinical outcome. AUC(0-∞) compresses information from a concentration-time profile into a single endpoint, while the geometric mean and geometric coefficient of variation describe the center and relative spread of those measurements. Reading these quantities correctly helps separate pharmacokinetic description from claims about clinical benefit or harm.

The record also highlights why comparator structure matters. LY2409021-only results are reported separately for Part A, Cohort 1; Part A, Cohort 2; and Part B participants using their Part A data. These are not interchangeable labels for one common control group. The combination conditions likewise belong to particular cohorts or parts. A simple comparison of arm summaries does not account for these design features or substitute for an analysis designed around them.

Another statistical lesson is the difference between variability and uncertainty in a contrast. The geometric coefficient of variation describes dispersion in the endpoint measurements. A confidence interval around a comparison would address uncertainty in the estimated contrast. The registry posts the former but not the latter, and it does not provide p-values. That distinction prevents a reader from interpreting an arm-level spread measure as evidence about the statistical precision of a treatment comparison.

Finally, this record shows the importance of transparent reporting. Information about the analysis model, missing observations, multiplicity, and sample-size justification is not reported in the registry record. Those details are necessary to evaluate how the estimates were obtained and how firmly they support conclusions beyond the descriptive summaries.

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 primary limitation for statistical interpretation is that the registry posts arm-level geometric means and geometric coefficients of variation without inferential comparison results. Confidence intervals, p-values, and contrast estimates are not reported, so the precision of differences cannot be evaluated from these summaries alone. The registry also gives no detailed statistical model or endpoint calculation procedure, including the handling of the extrapolated AUC portion.

Comparability is not straightforward across all listed arms. Gemfibrozil and its LY2409021-only observation are in Part A, Cohort 1; ketoconazole and its LY2409021-only observation are in Part A, Cohort 2. Part B enrolled only participants who had successfully completed Part A, and the Part B LY2409021-only observations are identified as data from Part A. The registry does not explain how these design features were handled in the analysis or whether adjustments were made.

The endpoint population includes only participants with at least 1 dose and evaluable AUC(0-∞) data. The record does not report participant flow for this endpoint, reasons for missing or nonevaluable measurements, or sensitivity analyses. It also does not report a multiplicity strategy, prespecified contrast hierarchy, or the trial’s sample-size justification. Although planning assumptions are described, the numerical illustrations are not part of the posted endpoint results.

The study enrolled healthy volunteers and is reported as Phase 1. The registry does not present clinical efficacy outcomes or secondary outcome results. The posted exposure summaries should therefore not be generalized to patient outcomes or treated as a complete account of safety or clinical effects.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: The Effect of Gemfibrozil, Ketoconazole and Clarithromycin on the Amount of LY2409021 in the Bloodstream — NCT01836198.
• Registry record last updated 2019-03-08; status last verified 2018-11.
• Results first posted 2019-03-08.
• 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

NCT01836198 is a completed, randomized, unmasked, parallel Phase 1 study in healthy volunteers, with enrollment of 30. It examined LY2409021 alone and with gemfibrozil, ketoconazole, or clarithromycin across Part A cohorts and Part B. The primary endpoint was AUC(0-∞) for LY2409021, reported as a geometric mean with a geometric coefficient of variation.

The registry posts endpoint summaries for six arm conditions. The largest posted geometric mean is 145000.0 ng*h/mL for LY2409021 plus ketoconazole in Part A, Cohort 2. No inferential comparison results or confidence intervals are posted. Serious adverse-event counts are zero in each displayed group; death counts are not reported in the registry record.

Important analysis details—including the statistical model, missing-data handling, multiplicity approach, and trial sample-size justification—are not reported in the registry record. The results are descriptive pharmacokinetic summaries and do not establish clinical outcomes.