A Study of LY2189265 Compared to Sitagliptin in Participants With Type 2 Diabetes Mellitus on Metformin: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study of LY2189265 Compared to Sitagliptin in Participants With Type 2 Diabetes Mellitus on Metformin (NCT00734474), a randomized phase 2, phase 3 study. The primary outcome is Glycosylated Hemoglobin (HbA1c) Change From Baseline.

2. Clinical Question

The study addressed glucose control in participants already receiving metformin. Its registry summary describes an adaptive dose-finding component alongside a Phase 3 efficacy study. The primary endpoint was change in glycosylated hemoglobin, or HbA1c, measured as a percentage of HbA1c. A negative change represents a lower follow-up value relative to baseline; comparisons should be interpreted using the stated group orientation and analysis method.

The primary endpoint focused on selected doses rather than every LY2189265 dose listed in the trial arms. The posted primary results include the 1.5 mg and 0.75 mg groups and sitagliptin. Other recorded outcomes include additional HbA1c time points, fasting blood glucose, fasting insulin, body weight, and other measures. Those outcomes address related but distinct questions and do not replace the primary comparison.

3. Trial Design

The registry describes a randomized, quadruple-masked, parallel-group Phase 2/3 study, sponsored by Eli Lilly and Company. It is marked completed and records enrollment of 1202 participants. The recorded start was 2008-08, and primary completion was 2011-06. The study was designed to evaluate once-weekly LY2189265 against sitagliptin in participants with type 2 diabetes mellitus on metformin.

The study included multiple LY2189265 dose groups, from 0.25 mg through 3.0 mg, as well as a sitagliptin group and placebo-containing arrangements. The arm descriptions specify subcutaneous weekly injections for LY2189265, oral daily tablets for sitagliptin, and metformin at least 1500 mg/day. Placebo injections or tablets were used in specified arms. The sitagliptin arm description lists treatment for 104 weeks; several LY2189265 arm descriptions also state treatment for up to 104 weeks. A placebo/sitagliptin arm description records placebo tablet use through 26 weeks followed by sitagliptin for 78 weeks.

The registry summary describes both dose finding and efficacy evaluation. The primary endpoint timeframe is baseline to 52 weeks, while some arm descriptions cover longer treatment periods. The registry record does not report the full allocation procedure, the masking roles, or the original statistical design assumptions. It also does not report a detailed account of how the adaptive dose-finding decisions were made. Those details should not be inferred from the trial label or the listed arms.

The primary analysis was described for randomized participants in the selected LY2189265 dose groups and sitagliptin who had evaluable HbA1c data. The record says missing postbaseline values were imputed using last observation carried forward. The posted primary analysis used ANCOVA, with country and treatment as fixed effects and baseline HbA1c as a covariate. Further details of the analysis population and missing-data rules are not reported in the registry record.

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.

Only the first 6 of 9 groups are shown.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: All randomized participants in the selected (1.5 mg, 0.75 mg LY2189265) and active comparator (Sitagliptin) arms who had evaluable HbA1c data. Last observation carried forward was used to impute missing postbaseline values. If there were no data after the date of randomization, the endpoint was considered missing.

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 randomized participants.

Only the first 6 of 10 groups are shown.

6. Primary Endpoint

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

The primary endpoint was change from baseline in glycosylated hemoglobin (HbA1c), measured as a percentage of HbA1c, at baseline and 52 weeks. HbA1c is used as a continuous measure of longer-term glucose exposure. Assessing change rather than only a follow-up value allows the analysis to describe how the measure shifted from each participant’s starting point over the specified period.

The registry reports least-squares means rather than only unadjusted arithmetic averages. In general, a least-squares mean is an estimate from a statistical model that accounts for terms included in that model. Here, ANCOVA included treatment and country as fixed effects and baseline HbA1c as a covariate. This approach accounts for the specified baseline measure and model terms when estimating the group results. The posted analysis also used last observation carried forward to impute missing postbaseline values.

The endpoint is relevant to the study’s stated efficacy question because the registry summary identifies change in HbA1c as its measure of glucose response. It does not, by itself, describe every aspect of diabetes control, the experience of participants, or longer-term clinical outcomes. A mean change also does not show the full spread of individual participant responses. The endpoint should therefore be read as a specific group-level measure at the recorded timeframe, rather than as a complete account of benefit or harm.

7. Results Posted in the Registry

• 0.75 mg LY2189265: mean -0.87 versus -1.10 on control; difference 0.23 (95% CI 0.06 to 0.40), standardized effect d = 0.22; z-test p = 0.007.
• Sitagliptin: mean -0.39 versus -1.10 on control; difference 0.71 (95% CI 0.54 to 0.88), standardized effect d = 0.68; z-test p = <0.001.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Change From Baseline in Glycosylated Hemoglobin (HbA1c) at the Dose Decision Point

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • 3.0 mg LY2189265: mean -1.09 versus -0.06 on control; difference -1.03 (95% CI -1.56 to -0.50), standardized effect d = -1.53; z-test p = <0.001.
  • 2.0 mg LY2189265: mean -1.25 versus -0.06 on control; difference -1.19 (95% CI -1.54 to -0.84), standardized effect d = -1.80; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for 3.0 mg LY2189265 than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for 2.0 mg LY2189265 than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Change From Baseline in Body Weight at Dose Decision Point

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • 3.0 mg LY2189265: mean -3.32 versus -0.56 on control; difference -2.76 (95% CI -4.62 to -0.90), standardized effect d = -1.20; z-test p = 0.004.
  • 2.0 mg LY2189265: mean -2.15 versus -0.56 on control; difference -1.59 (95% CI -2.50 to -0.68), standardized effect d = -0.87; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for 3.0 mg LY2189265 than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for 2.0 mg LY2189265 than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Change From Baseline in Pulse Rate at Dose Decision Point

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • 3.0 mg LY2189265: mean 6.63 versus 1.81 on control; difference 4.82 (95% CI 0.15 to 9.49), standardized effect d = 0.62; z-test p = 0.043.
  • 2.0 mg LY2189265: mean 3.43 versus 1.81 on control; difference 1.62 (95% CI -2.90 to 6.14), standardized effect d = 0.18; z-test p = 0.482.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for 3.0 mg LY2189265 than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between 2.0 mg LY2189265 and the control. The interval gives the range of effects that remain compatible with the data.

The registry lists a broad set of secondary outcomes, including additional HbA1c assessments, durability of HbA1c change, fasting blood glucose, fasting insulin, body weight, and durability of body-weight change. The posted secondary measures therefore cover related metabolic outcomes as well as changes over time. Some selected dose-decision results are available for HbA1c and body weight, while the registry records many more secondary outcome measures.

Across the posted dose-decision summaries, outcomes vary by measure and dose, and some groups have relatively small reported sample sizes. The body-weight and HbA1c summaries generally describe changes from baseline, while pulse rate is also posted as a dose-decision measure. These results add context to the primary HbA1c endpoint but do not answer the same question. They should be interpreted according to their own outcome definitions, timeframes, groups, and analysis methods rather than being combined into a single conclusion.

Secondary results carry less weight when considered alongside a primary endpoint because many outcomes and comparisons create more opportunities for chance findings. The registry record does not report a multiplicity adjustment or a complete hierarchy of secondary hypotheses. The displayed recalculated secondary comparisons are unadjusted and are not the sponsor’s reported ANCOVA results. They are useful as descriptive checks of the posted summaries, but they should not be treated as confirmatory evidence without the missing testing-plan details.

The secondary findings can help characterize the range of outcomes the study measured and identify questions for further analysis. They do not independently establish that a primary result is clinically meaningful, nor can they resolve limitations in the primary analysis. A full interpretation would require the prespecified secondary analysis plan and fuller information about missing data and multiplicity, which are not reported in the registry record.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Change From Baseline in Glycosylated Hemoglobin (HbA1c) at the Dose Decision Point
• Glycosylated Hemoglobin (HbA1c) Change From Baseline
• Durability of Change From Baseline in Glycosylated Hemoglobin (HbA1c)
• Fasting Blood Glucose Change From Baseline
• Fasting Insulin Change From Baseline
• Change From Baseline in Body Weight at Dose Decision Point
• Body Weight Change From Baseline
• Durability of Change From Baseline Body Weight

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is continuous, so the registry reports group means and mean contrasts. A mean summarizes the center of a distribution, but does not show its shape or how participant-level outcomes vary. The primary arm estimates are least-squares means, which come from the posted analysis model. The registry also lists standard errors for these means. In general, a standard error describes the sampling uncertainty of an estimated quantity; it is not the standard deviation of individual participant measurements.

What does ANCOVA contribute?

ANCOVA combines a comparison of treatment groups with adjustment for covariates. The registry says the primary ANCOVA included country and treatment as fixed effects and baseline HbA1c as a covariate. General statistical rationale: adjusting for a baseline measure can account for differences in starting values and can improve the precision of a group comparison when the model is appropriate. The adjusted estimate is conditional on the specified model and covariates, so it need not match a simple subtraction of reported arm means.

The sponsor’s analysis reports least-squares mean differences. The separate recalculations derive simple, unadjusted comparisons from posted summary statistics and approximate pooled standard deviations. They do not recreate the ANCOVA, its covariate adjustment, or its handling of the analysis population. As a result, the recalculated estimate and the sponsor’s estimate can differ in value or direction. Direction also depends on which group is subtracted from which: a contrast written as LY2189265 minus sitagliptin has the reverse sign from sitagliptin minus LY2189265.

What does last observation carried forward mean?

The registry states that last observation carried forward was used to impute missing postbaseline HbA1c values. In general, this method substitutes a participant’s last observed postbaseline value for a later missing value. It is a simple imputation approach, but it assumes that carrying the earlier measurement forward is an acceptable representation of the missing value. If outcomes continue to change after the last observation, that assumption may not reflect the unobserved course. The registry record does not report sensitivity analyses that test the impact of alternative missing-data assumptions.

How should the reported intervals and p-values be read?

A confidence interval expresses uncertainty around an estimate under the statistical model and its assumptions. A narrower interval indicates greater precision for the estimated contrast, while a wider interval indicates more uncertainty. A p-value measures how incompatible the observed data are with a specified null hypothesis under the analysis assumptions; it is not the probability that the null hypothesis is true. The registry reports 95% confidence intervals and p-values for the ANCOVA comparisons. The exact decision thresholds and the full planned testing strategy are not reported in the registry record.

The simple recalculations also include intervals and unadjusted p-values. Their approximate standard deviations are derived from the reported standard errors for the primary arm summaries. They are not participant-level analyses and do not account for the ANCOVA covariates. Because their assumptions and estimands differ from those of the sponsor’s model, these recalculations are best treated as a secondary check on the posted summary data, not as replacements for the registry analysis.

Why does multiplicity matter?

The registry lists multiple dose groups, outcomes, timeframes, and analyses. In general, when many hypotheses are tested, the chance of obtaining at least one apparently small p-value by chance can increase. Statistical plans may address this through prespecified testing sequences or adjustments, but the details of any multiplicity control are not reported in the registry record. Therefore, p-values for secondary outcomes and exploratory comparisons should be read with greater caution than a single prespecified primary test.

The study’s adaptive dose-finding description also makes the analysis plan important: dose selection and subsequent testing can affect how results should be interpreted. The registry summary identifies the study as adaptive but does not report enough detail to reconstruct the decision process or its statistical implications. The results can be described as posted, while uncertainty about those unreported methods remains explicit.

11. Confidence Intervals and Precision

The sponsor’s adjusted primary estimates have 95% confidence intervals of -0.87 to -0.55 for 1.5 mg LY2189265 versus sitagliptin and -0.63 to -0.31 for 0.75 mg LY2189265 versus sitagliptin. Both intervals are entirely on the negative side of zero in the reported contrast orientation. The sponsor also reports p < 0.001 for each comparison. These intervals quantify uncertainty in the model-based contrasts, given the analysis assumptions; they do not describe the range of individual participant changes.

The simple unadjusted recalculation for 1.5 mg LY2189265 versus sitagliptin is 0.71, with a 95% interval from 0.544 to 0.876. Its sign is reversed relative to the sponsor’s reported contrast because the group ordering differs. For 0.75 mg LY2189265 versus the 1.5 mg arm, the recalculated mean difference is 0.23, with a 95% interval from 0.064 to 0.396. That recalculation addresses a different pair of groups from the sponsor’s 0.75 mg versus sitagliptin comparison.

In general, interval width reflects precision: narrower intervals indicate a more precisely estimated contrast, while wider intervals indicate more uncertainty. The adjusted and recalculated intervals should not be compared as though they estimate exactly the same quantity. The sponsor’s values are model-adjusted; the recalculated values are unadjusted and rely on approximations from summary statistics. Neither interval establishes clinical importance, and neither gives the distribution of responses among individuals. The registry record does not report a threshold for a clinically meaningful HbA1c difference.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The planning figures are participants per arm for 80% power and participants per arm for 90% power. They are not total events: these outcomes are continuous, and the figures describe a two-arm planning illustration. The calculations assume a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. They use the observed effect or a stated fraction of that observed effect as the effect to detect.

The 80% power column gives the illustrative number of participants per arm needed for an 80% chance of detecting the specified effect if the planning assumptions hold. The 90% power column gives the corresponding per-arm number for a 90% chance. For the primary recalculation comparing 1.5 mg LY2189265 with sitagliptin, the table gives 35 participants per arm for 80% power and 46 participants per arm for 90% power when planning for 100% of the observed effect. For 50% of the observed effect, the corresponding figures are 138 and 184 participants per arm. This illustrates the general relationship: smaller effects require more participants to detect at the same power.

These are planning illustrations based on the observed data, not the trial’s original design calculation. The original design assumptions are not reported in the registry record. This is not observed power; it does not measure the probability that the trial’s actual result is true or reliable.

13. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = 0.230; pooled standard deviation σ = 1.038.
2. n = 2 × (1.96 + 0.842)² × 1.038² / 0.230² = 319.4, rounded up to 320 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
delta <- 0.23                      # observed difference in means
sigma <- 1.03752027553569                     # pooled standard deviation
n_per_arm <- function(power, frac = 1) {
  ceiling(round(2 * (z_a + qnorm(power))^2 * sigma^2 / (delta * frac)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (t-based, so very slightly larger)
cat("power.t.test:", round(power.t.test(delta = delta, sd = sigma, power = 0.80)$n, 1), "\n")

# Simulation: exact two-sample t-test using simulated means and variances
set.seed(734474)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 320 (80% power), 428 (90% power)
 75% of observed effect: n per arm = 568 (80% power), 761 (90% power)
 50% of observed effect: n per arm = 1,278 (80% power), 1,711 (90% power)
power.t.test: 320.4 
Simulated power at n = 320 per arm: 0.794 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

14. Analyses Reported by the Sponsor

The sponsor’s primary results are ANCOVA least-squares mean differences, adjusted for the model terms described in the registry. For 1.5 mg and 0.75 mg LY2189265 compared with sitagliptin, the reported differences are -0.71 and -0.47, respectively, each with a 95% confidence interval and p < 0.001. The registry lists these comparisons under both non-inferiority or equivalence and superiority or other analysis labels.

The recalculations use simple differences from posted arm summaries, not the sponsor’s ANCOVA model. For 1.5 mg versus sitagliptin, the recalculated estimate has the opposite sign from the sponsor’s result because the ordering of the groups is reversed. For 0.75 mg, the displayed recalculation compares that dose with 1.5 mg, not with sitagliptin; it is therefore not a recalculation of the sponsor’s reported -0.47 contrast.

The registry discrepancy flag is false. The difference between the calculations is explained by their methods and comparison orientation rather than a flagged mismatch. The sponsor’s adjusted analysis is the registry’s reported primary analysis; the simple recalculations are supplementary and should not be treated as substitutes for it. The registry record does not report the non-inferiority margin or full testing plan, so the formal decision framework cannot be reconstructed from the posted results alone.

15. Multiplicity, Missing Data and Analysis Populations

The primary analysis used last observation carried forward for missing postbaseline HbA1c values, according to the registry description. This is an explicit part of the reported endpoint analysis, but the record does not provide a full account of missingness by arm or explain the reasons values were missing. It also does not report sensitivity analyses using other assumptions about missing outcomes. In general, conclusions can depend on how missing data are handled, especially when missingness differs across groups or relates to participants’ outcomes.

The registry population description refers to all randomized participants in the selected LY2189265 and sitagliptin arms who had evaluable HbA1c data, with missing postbaseline values imputed by last observation carried forward. The remainder of the population rule is not reported in the registry record. The record therefore does not establish that the posted endpoint analysis includes every randomized participant under a complete intention-to-treat approach.

Multiplicity is also relevant because the study included multiple doses, outcomes, timeframes, and analyses. The record does not report whether p-values were adjusted across these tests, whether a fixed testing sequence was used, or how the adaptive dose-finding component affected confirmatory testing. The sponsor reports the primary comparisons under more than one analysis category, but the registry does not give enough detail to reconstruct the prespecified hypotheses, margins, or error-control strategy.

These omissions do not erase the results that are posted; they limit how fully the statistical evidence can be assessed. In general, an unadjusted p-value from one of many analyses gives less protection against chance findings than a result from a clearly prespecified and controlled testing plan. The registry record does not report the original sample-size justification, interim analyses, or the complete statistical analysis plan. No assumptions about those matters should be inferred.

16. Safety Results

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

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: gastroenteritis (6.7%), pneumonia (4.0%), thyroid cancer (3.3%), cervical dysplasia (1.1%), myocardial infarction (1.1%).

The registry’s serious adverse event summaries report, for each listed group, the number at risk and the number affected. For the 1.5 mg LY2189265 group, the serious-event summary lists 304 at risk and 36 affected; for 0.75 mg LY2189265, it lists 302 at risk and 23 affected; and for sitagliptin, it lists 315 at risk and 32 affected. Other dose and placebo/sitagliptin groups are also listed. These counts should be read within their named groups, not as direct evidence that one treatment is safer than another.

The registry also lists serious adverse-event terms with percentages, including gastroenteritis at 6.7%, pneumonia at 4.0%, thyroid cancer at 3.3%, cervical dysplasia at 1.1%, myocardial infarction at 1.1%, and blood calcitonin increased at 1.0%. The record does not make the basis for those term percentages sufficiently clear in the listed summary to support further interpretation. Death counts are not reported in the registry record.

Safety counts are descriptive. They do not provide adjusted comparisons, establish causation, or account for differences in exposure time. The listed group sizes and event counts should be interpreted alongside the trial’s specific reporting definitions and follow-up, details of which are not fully reported in the registry record.

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

The registry results show that the adjusted mean HbA1c changes in the selected LY2189265 groups were more negative than the adjusted mean change reported for sitagliptin at the primary 52-week assessment. The sponsor’s ANCOVA contrasts are -0.71 for 1.5 mg and -0.47 for 0.75 mg, each with a 95% confidence interval that remains below zero in the reported orientation. The p-values are reported as p < 0.001. These are findings about group-level changes under the stated model and analysis population.

They do not show that each participant experienced the group-average change, or that the observed differences necessarily matter to participants in practical terms. The registry record does not report a threshold for clinical importance. Nor do the HbA1c comparisons, by themselves, establish effects on complications, quality of life, or outcomes beyond the measures and timeframes listed.

The findings also do not settle every question about treatment safety. The adverse-event summaries are descriptive, and the posted data do not provide a complete adjusted safety comparison. The record’s missing details about multiplicity, adaptive decisions, analysis populations, and sensitivity to missing-data assumptions further limit a full assessment of the statistical evidence.

The recalculated estimates are simple checks based on posted summaries. They are not replacements for the sponsor’s adjusted analysis, and the group ordering and comparison pair must be checked before comparing their signs or magnitudes. Overall, the posted results describe specific comparisons in this completed randomized trial. They should not be generalized beyond the recorded population, interventions, endpoint, and follow-up without additional evidence.

18. Why This Trial Matters Statistically

Statistically, this trial is notable for combining an adaptive dose-finding study with a Phase 3 efficacy study and comparing several once-weekly LY2189265 doses with sitagliptin in participants taking metformin. Multiple dose groups can inform dose selection, while the primary endpoint provides a defined comparison at 52 weeks. The registry’s randomized, quadruple-masked, parallel-group design gives important context for interpreting the posted group estimates.

The record also illustrates why adjusted and unadjusted results should not be conflated. The sponsor’s primary ANCOVA accounts for baseline HbA1c, country, and treatment as specified in the model. Simple recalculations from summary statistics omit that adjustment and can use a different comparison orientation. Presenting both can help readers understand the reported summaries, but only if the estimands and group ordering remain clear.

Precision and planning are also distinct. The confidence intervals describe uncertainty in estimated contrasts, whereas the sample-size illustrations show how a hypothetical future study might be planned using stated assumptions and observed effects. Those illustrations are not evidence about the trial’s original design or its achieved power.

Finally, the record shows why completeness of statistical reporting matters. Details of multiplicity control, the adaptive decision process, the non-inferiority margin, and sensitivity analyses for missing data are not reported. Without them, readers can describe the posted results while recognizing limits on reconstructing the intended inferential framework.

19. 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 leaves several trial-specific issues unresolved. Although it describes the study as adaptive and includes multiple dose groups, it does not explain the dose-selection decisions or the statistical consequences of adaptation. The original sample-size justification and any interim analyses are not reported. The sample-size illustrations based on observed effects cannot fill that gap.

The formal comparison framework is also incomplete in the record. The primary comparisons are labeled under both non-inferiority or equivalence and superiority or other analyses, but the non-inferiority margin, prespecified testing sequence, and multiplicity-control strategy are not reported. This prevents independent reconstruction of the formal decision rules.

The primary population is described as randomized participants in selected arms who had evaluable HbA1c data, with last observation carried forward for missing postbaseline values. The full analysis-population rule, missingness by group, and sensitivity analyses are not reported. The posted model-adjusted results therefore cannot be checked against alternative missing-data assumptions using the registry summaries alone.

The available recalculations rely on aggregate summaries rather than participant-level data. They cannot reproduce the sponsor’s ANCOVA or assess model fit, covariate balance, or individual outcome distributions. The arm-level estimates and safety counts also should not be mistaken for complete descriptions of participant-level variation or exposure-adjusted risk. Finally, the registry summaries do not provide enough information to assess the clinical importance threshold for HbA1c differences or to make a comprehensive comparison of safety.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: A Study of LY2189265 Compared to Sitagliptin in Participants With Type 2 Diabetes Mellitus on Metformin — NCT00734474.
• Registry record last updated 2015-04-20; status last verified 2015-03.
• Results first posted 2015-01-28.
• 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.

23. Record Summary

NCT00734474 is a completed randomized, quadruple-masked, parallel-group Phase 2/3 study sponsored by Eli Lilly and Company. The registry records enrollment of 1202 participants with type 2 diabetes mellitus on metformin. The study compared once-weekly LY2189265 across multiple doses with sitagliptin and placebo-containing arrangements.

The primary endpoint was least-squares mean change in HbA1c from baseline to 52 weeks. The registry reports ANCOVA comparisons of 1.5 mg and 0.75 mg LY2189265 with sitagliptin, using baseline HbA1c as a covariate and last observation carried forward for missing postbaseline values. Additional outcomes include glycemic measures, body weight, and other measures.

Interpretation is limited by unreported details about multiplicity, the adaptive dose-selection process, the formal non-inferiority framework, and sensitivity analyses for missing data. Simple recalculations are unadjusted summaries and should remain distinct from the sponsor’s model-based primary results.