A Study of LY2189265 in Japanese Patients With Type 2 Diabetes: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study of LY2189265 in Japanese Patients With Type 2 Diabetes (NCT01001104), a randomized phase 2 study. The primary outcome is Change From Baseline in Glycosylated Hemoglobin (HbA1c) at 12 Weeks.

2. Clinical Question

The central statistical question is about the size and precision of the estimated HbA1c changes, not simply whether a test crossed a significance threshold. The registry reports mean changes for each group with 95% confidence intervals, as well as mixed-model comparisons of each LY2189265 dose with placebo. Taken together, these results describe the observed glycemic pattern over the 12-week treatment period.

The study's stated purpose was to assess dose-response characteristics in Japanese patients with type 2 diabetes taking LY2189265 monotherapy. The record also lists secondary measures that cover additional aspects of glycemic control, metabolic measures, body weight, and steady-state drug concentrations. Those measures can add context, but they do not replace the primary outcome or resolve questions beyond the outcomes and follow-up reported in the registry record.

3. Trial Design

This was a phase 2, randomized, triple-masked, parallel-group study sponsored by Eli Lilly and Company. It evaluated LY2189265 and placebo in Japanese patients with type 2 diabetes. The registry lists three LY2189265 dose groups—0.25 mg, 0.5 mg, and 0.75 mg—and a placebo group. The study's brief summary describes LY2189265 as monotherapy.

The registry reports enrollment of 145. For the primary outcome, the stated analysis population was the full analysis set: all randomized participants who received at least 1 dose of study drug. The primary outcome group counts were 34.0 for 0.75 mg, 37.0 for 0.5 mg, 36.0 for 0.25 mg, and 36.0 for placebo. These are outcome-analysis counts, not a replacement for the registry's enrollment figure. The record does not give further detail about allocation ratios or the randomization procedure.

The primary endpoint was change in HbA1c from baseline to 12 weeks. The registry gives the primary outcome timeframe as baseline and 12 weeks, and the study's listed dates are 2009-10 for start and 2010-12 for primary completion. The study status is completed. No additional follow-up duration is reported in the registry record.

Other listed outcomes include the proportion of participants reaching HbA1c thresholds, fasting and mean daily blood glucose, body weight, insulin sensitivity, beta-cell function, and steady-state LY2189265 concentrations. The registry record does not give a detailed protocol narrative, planned sample-size justification, or further operational details about conduct beyond the design and analysis descriptions.

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.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Full analysis set: All randomized participants who received at least 1 dose of the study drug.

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) at 12 Weeks, measured over Baseline, 12 weeks.

The primary endpoint measures change in glycosylated hemoglobin (HbA1c) from baseline to 12 weeks, calculated as the week 12 value minus the baseline value. HbA1c is a measure of glycemic control, and a change score expresses how the measure shifted during the treatment period. In this study, a negative change means the week 12 value was lower than baseline; the endpoint itself does not establish why a change occurred.

The registry reports least-squares means rather than simple unadjusted group averages. These are model-based estimates from a mixed model repeated measures analysis. The model included dose, pre-study therapy, baseline body mass index group, baseline value, visit, and the interaction between dose and visit, with participant treated as a random effect. This approach uses the repeated observations described by the model while accounting for the listed covariates and the within-participant structure.

Using HbA1c as the primary measure gives the study a specified, continuous outcome for comparing the dose groups with placebo over the stated treatment period. The reported 95% confidence intervals describe uncertainty around each estimated mean change. They are not ranges containing 95% of individual participants' changes, and they do not by themselves describe the full distribution of participant-level responses.

The registry defines the endpoint and analysis but does not report an interpretation threshold for a clinically important change. The statistical results therefore characterize estimated changes and their precision; they do not, by themselves, determine the practical importance of those changes.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does the mixed model estimate?

The primary analysis is described as a mixed model repeated measures (MMRM) analysis. It estimates mean change in HbA1c while including dose, pre-study therapy, baseline body mass index group, baseline value, visit, and dose-by-visit interaction. Participant is treated as a random effect. In general, a mixed model can represent the correlation among repeated measurements from the same participant, rather than treating each visit measurement as unrelated. The registry does not report additional details about model fitting, covariance structure, or how missing observations were handled.

The reported primary values are least-squares means. In general, a least-squares mean is a model-based estimate that accounts for the factors included in the analysis; it need not equal the raw arithmetic mean of the observed outcomes. This matters because the primary group estimates and the between-group contrasts come from the described model. They should not be interpreted as simple subtractions of raw averages unless the analysis specifically supports that interpretation.

How should confidence intervals be read?

A confidence interval describes uncertainty around an estimated quantity under the analysis approach. The registry reports 95% confidence intervals for each group's primary mean change. In general, a narrower interval indicates greater precision than a wider interval, while neither interval width nor the point estimate alone establishes clinical importance. A confidence interval for a group mean is not an interval intended to contain 95% of individual outcomes.

The posted primary results do not include confidence intervals for the least-squares mean differences against placebo. This means the precision of the primary contrasts is less fully described than the precision of the group-specific mean changes. A p-value for a contrast does not supply the missing interval or indicate the plausible range of contrast estimates.

What do the reported tests address?

The registry reports an overall mixed models analysis across the four primary groups, and pairwise mixed-model comparisons of each active dose with placebo. The overall test and the pairwise tests address different questions: an overall test assesses evidence of differences across the group set, while a pairwise contrast focuses on a particular dose-versus-placebo comparison. The three reported pairwise p-values are each p < 0.001. The registry does not report a multiplicity adjustment for these comparisons, so the record does not establish whether or how the family of tests was controlled.

For binary secondary outcomes, the registry also reports Cochran-Armitage tests across groups and Fisher exact tests for some specified pairwise comparisons. In general, a Cochran-Armitage test can assess an ordered trend across groups when an ordering is specified. Fisher's exact test evaluates association in a two-group table and can be useful when counts are sparse. These tests answer distinct questions and should not be treated as interchangeable. The record does not describe a multiplicity strategy for the secondary tests.

How should the binary re-calculations be understood?

Some secondary comparisons include recalculated risk differences and confidence intervals from the posted counts, along with unadjusted p-values. A risk difference expresses the difference between two observed proportions. In general, the confidence interval describes uncertainty around that difference; if it spans zero, the data are compatible with no difference as well as with values on either side, under the calculation used. These recalculations are simple, unadjusted comparisons and do not replace the sponsor's own analyses. They should not be interpreted as adjusted estimates or as evidence that other sources of uncertainty have been resolved.

The registry's reported analyses and the simple re-calculations may differ in question, adjustment, or method. In particular, a model-based result that accounts for listed covariates is not equivalent to a direct calculation from group counts. The primary endpoint has model-based sponsor results; no recalculated primary contrast is posted in the registry record.

What do the planning illustrations mean?

Planning illustrations use the observed effect as a starting point and consider detecting a stated fraction of that effect. The stated assumptions are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. In general, a smaller target effect requires more participants for a specified power when other assumptions are held constant. These are planning illustrations, not observed power, and they are not the trial's own sample-size justification. The sample-size justification is not reported in the registry record.

More generally, a sample-size calculation is conditional on its assumptions about the outcome, effect size, allocation, significance level, and analysis. The observed effect in one trial may be an imprecise estimate of the effect expected in a future study. Thus, the illustrations can help show how planning changes with the assumed effect, but they do not demonstrate that the completed trial had a particular probability of detecting its observed result.

What remains unspecified?

The record does not describe the model's missing-data assumptions, imputation procedures, any interim analyses, or the full statistical analysis plan. It identifies the full analysis set for the primary outcome and states which participants qualify for it, but does not give further details about outcome availability or exclusions. Those omissions limit independent assessment of sensitivity to missing data and of the prespecification of analyses. They do not negate the posted results, but they define the limits of what can be concluded from the registry record.

10. Analyses Reported by the Sponsor

The registry reports the sponsor's primary analysis as a mixed models analysis, with least-squares mean changes and dose-versus-placebo least-squares mean differences. The posted contrasts are -0.72 for 0.25 mg, -0.97 for 0.5 mg, and -1.17 for 0.75 mg, each with p < 0.001. These are model-based estimates consistent with the stated MMRM approach, which includes baseline and treatment-related factors and treats participant as a random effect.

No recalculated primary contrast is posted in the registry record. A primary re-calculation cannot be inferred from the arm-level means alone because the sponsor's contrast comes from the specified model, and the available summary values do not reproduce all model information. The sponsor analysis therefore remains the primary comparison to interpret. The record does not flag a discrepancy between the posted primary analysis and a recalculation.

For certain secondary binary outcomes, simple unadjusted risk-difference calculations are available from the posted group counts. Those calculations are not the same as the registry's reported Cochran-Armitage or Fisher exact analyses and should be treated as supplementary descriptions rather than replacements for the sponsor's analysis. Where a discrepancy is flagged, the sponsor's own analysis takes precedence; no such discrepancy is indicated in this record.

11. Multiplicity, Missing Data and Analysis Populations

The registry reports several primary tests: an overall mixed models analysis across four groups and three dose-versus-placebo comparisons. It also lists multiple secondary outcomes and reports more than one kind of statistical test for some of them. The multiplicity strategy for these tests is not reported in the registry record. In particular, the record does not state whether a hierarchy, adjustment, or other procedure was used to control the chance of false-positive findings across the outcome and comparison set.

Without that information, the reported p-values should be interpreted as results for their stated tests, not automatically as multiplicity-adjusted evidence across all analyses. This limitation is especially relevant to secondary outcomes, where the number of possible findings is broader. The primary endpoint remains the clearest prespecified focus in the record, but the analysis plan details needed to confirm how its multiple comparisons were handled are not reported.

The registry describes the primary full analysis set as all randomized participants who received at least 1 dose of study drug. That definition clarifies who qualified for the stated primary analysis. It does not, on its own, establish how many participants had an observed HbA1c value at each visit, how missing observations were handled, or whether other analysis populations were used. The MMRM description identifies the repeated-measures model and its listed covariates, but missing-data assumptions, imputation methods, sensitivity analyses, and reasons for missing data are not reported in the registry record.

Outcome-analysis counts vary across posted measures, so the denominator for one outcome should not be assumed to apply to another. The record does not provide enough detail to determine whether any differences in counts reflect missing outcome data, analysis-population definitions, or another reason. The distinction between enrollment, the stated primary analysis population, and outcome-specific counts matters when judging how representative each estimate is of all randomized participants.

The registry does not report whether interim analyses were planned or conducted, nor does it give a full statistical analysis plan. These details are important for evaluating prespecification and the handling of repeated opportunities to examine outcomes. Their absence is a limitation in the available record, not evidence that a particular procedure was or was not used.

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 2% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: pancreatic carcinoma (2.9%), viith nerve paralysis (2.8%).

The safety figures are descriptive counts of participants affected, with the number at risk reported for each group. For serious adverse events, the record reports 0 affected in the placebo group, 1 in the 0.25 mg group, 0 in the 0.5 mg group, and 1 in the 0.75 mg group. The record also lists Pancreatic carcinoma at 2.9% and VIIth nerve paralysis at 2.8% among the top serious-event terms. The relationship between those listed terms and individual treatment groups is not detailed in the available summary.

These counts should not be treated as estimates of comparative risk with strong precision: the groups are small, and the registry record does not provide confidence intervals or a comparative safety analysis for these figures. The listed terms do not establish that an event was caused by study treatment. Deaths are not reported in the registry record.

Other adverse events are also reported descriptively: 16 participants in placebo, 17 in the 0.25 mg group, 18 in the 0.5 mg group, and 15 in the 0.75 mg group were affected. These counts are not a substitute for event-level detail, severity, timing, or causal assessment, which are not reported in the registry record. The figures describe what was recorded, without establishing a safety difference between groups.

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

The primary results show model-based HbA1c reductions from baseline in all four groups over the 12-week period, with larger estimated reductions across the higher LY2189265 doses. The registry's pairwise mixed-model comparisons with placebo are reported as statistically significant at p < 0.001 for each active dose. Those results describe the posted data under the stated model and support a dose-ordered pattern in this trial.

Statistical evidence is not the same as a judgment about clinical importance. The registry does not report a threshold for a clinically meaningful change, and the confidence intervals describe uncertainty around estimated means rather than individual responses. The contrast confidence intervals are not posted, limiting the direct assessment of precision for the dose-versus-placebo effects.

The findings apply to the population, treatment conditions, analysis population, and follow-up described in the record. They do not establish whether similar results would occur in other populations or settings, whether the observed pattern persists beyond 12 weeks, or how treatment compares with therapies not studied as a comparator. The trial compares three doses of LY2189265 with placebo; it does not provide a comparison with another active treatment.

The results also do not answer every question about benefit and harm. The record includes some secondary outcomes and descriptive safety counts, but several listed secondary outcomes have no numerical results reported in the registry record, and the safety summary does not provide detailed event-level comparative analysis. A small p-value for the primary endpoint does not resolve uncertainties about missing data, multiplicity, longer-term effects, or outcomes that were not posted.

Accordingly, the appropriate interpretation is statistical and bounded: the trial's posted primary analysis shows a dose-related pattern of estimated HbA1c change over the stated period, with reported dose-versus-placebo tests. It is not, by itself, a complete basis for conclusions about broader treatment effects, long-term outcomes, or individual patient decisions.

14. Why This Trial Matters Statistically

This study is statistically informative because it evaluates several LY2189265 doses against placebo and reports a continuous primary outcome at a defined follow-up. A multi-dose design can make a dose-ordered pattern visible rather than limiting the question to a single active dose. Here, the model-based mean changes become more negative across the three active doses, and the registry reports dose-versus-placebo contrasts for each of them.

The primary analysis also illustrates why the reported estimand and model matter. The registry presents least-squares means from a repeated-measures model that includes baseline and other specified factors. Those estimates answer a model-based question and are not interchangeable with unadjusted arithmetic averages. Clear descriptions of the outcome, analysis population, and covariates help readers understand what the primary estimates represent.

At the same time, the record makes visible common limits of registry results. Group-level confidence intervals are reported for the primary means, but not for the primary contrasts. The record does not state the multiplicity approach, missing-data procedures, or sample-size justification. Secondary outcomes broaden the assessment, yet many listed outcomes have no numerical result in the record, and the results that are present do not replace the primary analysis.

For clinical researchers and biostatisticians, the main lesson is to distinguish a coherent statistical signal from the full body of evidence needed to judge a treatment. The dose pattern and model-based comparisons are useful findings, but precision, analysis assumptions, outcome availability, and the study's 12-week scope remain central to interpretation. The study is therefore valuable as a dose-response assessment while also illustrating why a registry summary should be read with attention to what it reports and what remains unspecified.

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.

Several limitations are specific to the available record. Although the primary analysis is described as MMRM, further model details are not reported: the covariance structure, estimation details, diagnostic checks, and sensitivity analyses are absent. The registry does not describe how missing outcomes were handled or what assumptions support the repeated-measures analysis. These omissions prevent a full independent assessment of robustness to missing data or model specification.

The primary results include confidence intervals for group means but not for the reported least-squares mean differences. This makes it harder to assess the precision of the comparisons central to the dose-versus-placebo question. The p-values indicate evidence under the reported model, but do not supply a plausible range for each contrast.

The record does not report a multiplicity strategy despite the overall test, several pairwise comparisons, and a broad set of secondary outcomes. Nor does it report a sample-size justification, interim-analysis information, or a complete statistical analysis plan. The planning illustrations based on observed effects are not evidence of the original design assumptions or of the completed trial's observed power.

The primary endpoint covers 12 weeks, and the available record does not establish durability beyond that period. The study concerns Japanese patients with type 2 diabetes and does not report results for other populations. Numerical results for several listed secondary outcomes are not reported in the registry record, limiting assessment of whether changes in other metabolic measures align with the primary outcome.

Safety reporting is also limited in the summary: serious adverse-event counts and selected terms are listed, but detailed event-level information, comparative safety estimates, and deaths are not reported in the registry record. These limitations concern the completeness and interpretability of the posted record; they do not establish that unreported procedures or data were absent from the full study documentation.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: A Study of LY2189265 in Japanese Patients With Type 2 Diabetes — NCT01001104.
• Registry record last updated 2015-09-28; status last verified 2015-08.
• Results first posted 2015-09-28.
• 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, triple-masked, parallel phase 2 trial of LY2189265 monotherapy in Japanese patients with type 2 diabetes. It lists placebo and three active dose groups, and enrollment of 145.

The primary endpoint was change in HbA1c from baseline to 12 weeks. The stated full analysis set included randomized participants who received at least 1 dose of study drug. A mixed model repeated measures analysis reported dose-ordered least-squares mean changes and statistically significant comparisons of each active dose with placebo.

Secondary outcomes included HbA1c thresholds and other glycemic, metabolic, weight, and drug-concentration measures. Numerical results for several listed outcomes are not reported in the registry record. Serious adverse-event counts are descriptive, with limited comparative detail. Multiplicity procedures, missing-data handling, sample-size justification, and further statistical-plan details are not reported in the registry record.