GLP-1 Receptor Agonist Lixisenatide in Patients With Type 2 Diabetes for Glycemic Control and Safety Evaluation, on Top of Basal Insulin: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for GLP-1 Receptor Agonist Lixisenatide in Patients With Type 2 Diabetes for Glycemic Control and Safety Evaluation, on Top of Basal Insulin (NCT00715624), a randomized phase 3 study. The primary outcome is Absolute Change From Baseline in Glycosylated Hemoglobin (HbA1c) at Week 24.

2. Clinical Question

The clinical question concerns adding lixisenatide to basal insulin for people with type 2 diabetes whose glycemic control was insufficient with basal insulin. The comparison was with placebo added to basal insulin, with or without metformin. The primary focus was the change in glycosylated hemoglobin (HbA1c) from baseline to Week 24. HbA1c is a measure of average blood glucose over time, so its change offers a way to compare glycemic control between the randomized groups over the main treatment period.

The record frames the trial as an evaluation of efficacy and safety, and lists further outcomes concerning glucose measures, body weight, insulin dose, rescue therapy, tolerability, pharmacokinetics, and antibodies. The primary question is narrower: it concerns the between-group contrast in HbA1c change during the main treatment period, not every outcome or the extension period.

3. Trial Design

This was a randomized, double-blind, placebo-controlled, two-arm, parallel-group, multicenter Phase 3 study. The registry record lists quadruple masking. Its main treatment period lasted 24 weeks and was followed by an extension. Enrollment was recorded as 496. The study status is completed; the recorded start was 2008-07 and primary completion was 2011-02.

The trial compared lixisenatide with volume-matching placebo as an add-on to basal insulin. Basal insulin and metformin appear in the intervention record as background treatments; the study description specifies basal insulin with or without metformin. The lixisenatide arm used a stepwise initiation regimen: 10 microgram once daily for one week, then 15 microgram once daily for one week, followed by 20 microgram once daily through the end of treatment. The placebo description gives a matching stepwise schedule.

Random allocation supports a comparison between assigned groups, while masking is intended to limit the influence of treatment knowledge on participants’ experiences and trial assessments. These design features do not, by themselves, establish that every possible source of bias was eliminated. The registry record does not report the allocation ratio or detailed procedures for sequence generation and concealment, so those details are not reported in the registry record.

The primary endpoint was assessed at Week 24, within the main treatment period. The endpoint description defines the efficacy observation window in relation to dosing, the Week 24 visit, and rescue therapy. An extension followed the main period, but extension results are not reported in the registry record. The primary completion date does not describe the extension’s duration or its findings.

The record identifies an mITT population for the primary outcome and describes the analysis method as ANCOVA. It does not report the trial’s original sample-size justification, detailed model specification, or a full statistical analysis plan. Those design details 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.

Before assignment: A total of 879 patients were screened of which 383 (43.6%) were screen failures; main reason for screen failure was glycosylated hemoglobin (HbA1c) values being out of the defined protocol range (greater than or equal to 7% and less than or equal to 10%). A total of 496 patients were randomized.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: mITT population:all randomized patients who received at least 1 dose;had baseline,at least 1 post-baseline efficacy assessment, irrespective of compliance with study protocol/procedures. LOCF was used. Number of patients analyzed=patients with baseline and at least 1 post-baseline HbA1c assessment during on-treatment period.

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: Safety population included all randomized patients who were exposed to at least 1 dose of study drug, regardless of the amount of treatment administered.

6. Primary Endpoint

The registry lists the primary outcome as Absolute Change From Baseline in Glycosylated Hemoglobin (HbA1c) at Week 24, measured over Baseline, Week 24.

The primary endpoint is absolute change in HbA1c from baseline to Week 24, measured in percentage of hemoglobin. The registry defines change as the Week 24 HbA1c value minus the baseline value. A negative change therefore represents a lower measured HbA1c at Week 24 than at baseline. The endpoint summary reports least-squares means and standard errors for each group.

HbA1c is used as a measure of glycemic control because it reflects glucose exposure over time rather than a single glucose reading. Comparing its change between randomized groups addresses whether their average HbA1c trajectories differed during the main treatment period. The comparison is more informative than interpreting either group’s change in isolation: change in the placebo group also captures changes that can occur over time under trial conditions and background treatment.

The primary analysis population was described as randomized patients who received at least one dose and had a baseline and at least one post-baseline efficacy assessment. The record says last observation carried forward (LOCF) was used. For the efficacy variable, the on-treatment period extended from the first dose through three days after the last dose, subject to the Week 24 visit or Day 169 and preceding rescue therapy. Thus the endpoint is tied to a specified treatment window and analysis population; it is not a direct summary of outcomes in every randomized patient at every point in the extension.

The primary endpoint is a continuous measure. Its group estimates and the between-group estimate describe mean changes, not the proportion of individuals who reached a particular HbA1c threshold. Threshold outcomes are listed among the secondary outcomes, but their numerical results are not reported in the registry record.

7. Results Posted in the Registry

• Lixisenatide: mean -0.74 versus -0.38 on control; difference -0.36 (95% CI -0.63 to -0.09), standardized effect d = -0.24; z-test p = 0.010.

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 2-Hour Postprandial Plasma Glucose (PPG) at Week 24

• 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 errors.
• Result:
  • Lixisenatide: mean -5.54 versus -1.72 on control; difference -3.82 (95% CI -5.22 to -2.42), standardized effect d = -0.56; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Lixisenatide than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Change From Baseline in Average 7-Point Self Monitored Plasma Glucose (SMPG) Profiles at Week 24

• 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 errors.
• Result:
  • Lixisenatide: mean -1.49 versus -0.61 on control; difference -0.88 (95% CI -1.49 to -0.27), standardized effect d = -0.27; z-test p = 0.005.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Lixisenatide than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Change From Baseline in Fasting Plasma Glucose (FPG) at Week 24

• 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 errors.
• Result:
  • Lixisenatide: mean -0.63 versus -0.55 on control; difference -0.08 (95% CI -0.80 to 0.64), standardized effect d = -0.02; z-test p = 0.827.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Lixisenatide and the control. The interval gives the range of effects that remain compatible with the data.

The numerical secondary results cover postprandial glucose, average seven-point self-monitored glucose profiles, and fasting plasma glucose. Across these measures, the pattern is not uniform: the recalculated estimates for the first two favor a larger mean reduction in the lixisenatide group, while the fasting measure gives little separation and substantial uncertainty around its estimate. This broad pattern can be considered alongside the primary HbA1c result because all concern glycemic measures, but the secondary endpoints are not interchangeable with HbA1c and do not replace the primary comparison.

These secondary comparisons are simple unadjusted recalculations from summary statistics. The registry record has no sponsor-reported analyses for these three outcomes in the posted analysis details. Their intervals and p-values therefore do not have the same status as the sponsor’s reported primary ANCOVA result. In particular, a collection of unadjusted tests raises a multiplicity concern: the more outcomes examined, the more opportunity there is for some apparently notable findings to occur by chance. An adjustment or hierarchical testing strategy is not reported in the registry record.

The listed secondary outcomes also include body weight, insulin dose, HbA1c thresholds, and rescue therapy. Numerical results for those outcomes are not reported in the registry record. Consequently, the overall secondary-outcome picture is incomplete; no direction or magnitude should be inferred for outcomes without posted numerical estimates.

Secondary findings can add context, suggest whether different measures tell a broadly coherent story, and identify questions for further study. Their evidential weight depends on prespecification, multiplicity control, analysis methods, and completeness of reporting. Those details are not fully available in the registry record. The safest reading is to treat the reported secondary pattern as supportive descriptive evidence, not as a set of definitive independent claims.

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

• Change From Baseline in 2-Hour Postprandial Plasma Glucose (PPG) at Week 24
• Change From Baseline in Average 7-Point Self Monitored Plasma Glucose (SMPG) Profiles at Week 24
• Change From Baseline in Fasting Plasma Glucose (FPG) at Week 24
• Change From Baseline in Body Weight at Week 24
• Change From Baseline in Total Insulin Dose at Week 24
• Percentage of Patients With Glycosylated Hemoglobin (HbA1c) Level Less Than 7% at Week 24
• Percentage of Patients With Glycosylated Hemoglobin (HbA1c) Level Less Than or Equal to 6.5% at Week 24
• Percentage of Patients Requiring Rescue Therapy During Main 24-Week Period

9. Statistical Interpretation

10. Statistical Methods Explained

The primary result is presented as a comparison of continuous outcomes. A continuous measure can take a range of values, such as HbA1c change, rather than being only a yes-or-no event. A group mean summarizes the average observed change, while a between-group mean difference expresses how far apart the group averages are. Because both groups can change during a trial, the between-group contrast is the central comparison rather than either group’s change alone.

What does ANCOVA contribute?

The registry identifies ANCOVA as the sponsor’s method for the primary comparison and reports a least-squares mean difference. In general, analysis of covariance compares group outcomes while accounting for specified covariates. This can improve precision when the covariates are relevant and the model is appropriate. The registry record does not identify the covariates, model terms, or diagnostic checks used for the ANCOVA, so those details are not reported in the registry record. The reported adjusted result should be understood as the sponsor’s model-based comparison, not as a raw subtraction of the posted group summaries.

The least-squares means are model-based group estimates. Their interpretation depends on the fitted model and its covariates. The record’s primary endpoint summary gives the least-squares mean changes and standard errors, while the sponsor analysis gives the LS mean difference, confidence interval, and p-value. These quantities work together: the group estimates give the direction and scale of change, and the contrast estimates the difference between groups.

How does the recalculation differ?

The recalculated comparison is an unadjusted check based on the posted summary statistics. It does not reproduce the sponsor’s ANCOVA because the registry record does not provide the full model specification or participant-level data. The recalculation uses the two group means and their reported standard errors to estimate a difference and interval. Its pooled standard deviation is approximate, derived from summary information rather than individual observations. Such an exercise can help readers assess whether the broad direction and scale are coherent, but it cannot recover covariate adjustment, missing-data handling beyond what is encoded in the summaries, or other details of the sponsor model.

Here the recalculated estimate is -0.36, with an interval from -0.634 to -0.086 and an unadjusted p-value of 0.01. The sponsor’s ANCOVA estimate is -0.36, with a 95% confidence interval from -0.550 to -0.174 and p < 0.001. The discrepancy flag means the sponsor analysis takes precedence. Differences between an adjusted model and an unadjusted summary-based calculation are possible because they do not necessarily estimate uncertainty in the same way or adjust for the same information.

What does a standard error describe?

A standard error describes the estimated variability of a statistic across repeated samples under a statistical model. It is not the spread of individual patient outcomes. A smaller standard error generally indicates a more precise estimate, although precision also depends on the quantity being estimated and the model assumptions. The primary group estimates are reported with standard errors, and the recalculation uses those summaries.

When a continuous outcome has greater variability among participants, a group mean is generally estimated less precisely for a given amount of information. In general, this is one reason variability matters when interpreting a mean difference and when planning a study. A pooled standard deviation can express variability on the outcome scale, but the recalculated pooled standard deviation is approximate and should not be confused with the sponsor’s adjusted analysis.

How should p-values be read?

A p-value describes how unusual data at least as incompatible with a specified null model would be if that null model and the analysis assumptions held. It is not the probability that a treatment effect is real, nor does it describe the probability of a patient’s outcome. The sponsor reports p < 0.001 for the primary ANCOVA analysis. The recalculated unadjusted result has p = 0.01. Secondary p-values are unadjusted recalculations and should be interpreted with additional caution because several outcomes are considered.

Statistical significance alone does not state whether an effect is clinically important. That judgment requires context about the measure, its meaningful change, benefits and harms, and the population to which the result applies. The registry record does not provide a threshold for a clinically important HbA1c difference, so such a threshold is not reported in the registry record.

Why can over-dispersion matter?

For outcomes that are counts or proportions, a model may underestimate uncertainty if observations vary more than its assumed distribution allows. This phenomenon is generally called over-dispersion. It matters because standard errors that are too small can make confidence intervals too narrow and p-values too small. The primary endpoint is continuous and is analyzed by ANCOVA, so over-dispersion is not the central issue for this primary comparison. The registry record does not describe an over-dispersion assessment or a count model.

More broadly, the analysis depends on the population definition, outcome timing, missing-data handling, and assumptions of the fitted method. The registry states that LOCF was used for the primary efficacy analysis and specifies an mITT population. It does not report sensitivity analyses for missing data, the complete ANCOVA specification, or diagnostics. Those details are not reported in the registry record, limiting independent reproduction of the sponsor’s analysis.

11. Confidence Intervals and Precision

A confidence interval gives a range of values compatible with the data and the statistical procedure, under its assumptions. It conveys uncertainty about an estimated difference rather than identifying the one true value. A narrower interval indicates greater statistical precision; a wider interval allows a broader range of possible effects. Precision is distinct from clinical importance.

For the sponsor’s primary ANCOVA result, the reported LS mean difference is -0.36, with a 95% confidence interval from -0.550 to -0.174. The interval lies below zero, supporting the direction of a greater mean HbA1c reduction in the lixisenatide group in this analysis. The interval does not mean that every participant had a reduction, and it does not provide a range for individual responses. It describes uncertainty around the estimated group contrast.

The recalculated unadjusted interval runs from -0.634 to -0.086. It also lies below zero, but it is wider than the sponsor’s reported interval. This is a different calculation, not a correction to the sponsor result. The discrepancy flag means the sponsor’s own analysis takes precedence; the recalculated interval is best read as a summary-statistic check with approximate variability, not as a competing adjusted analysis.

The secondary recalculated intervals have different degrees of precision. The postprandial glucose and seven-point profile intervals do not include zero, while the fasting plasma glucose interval extends from -0.795 to 0.635 and includes zero. An interval including zero indicates that the data and calculation do not sharply distinguish the estimated difference from no difference at the stated confidence level. It does not establish equivalence or prove there is no effect.

Confidence intervals do not account for every source of uncertainty. The registry’s method details, the outcome-specific analysis populations, and the handling of missing observations affect how far these intervals can be independently assessed. Statistical intervals also do not address multiplicity unless that is built into the procedure; adjustment for multiple testing is not reported in the registry record.

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 illustrations for continuous outcomes and are expressed as participants per arm, not total events. The registry’s planning assumptions are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. The effect to detect is set to a stated fraction of the observed effect. The original trial’s own sample-size assumptions and justification are not reported in the registry record.

The 80% power column indicates the number of participants per arm in a hypothetical planning exercise intended to give an 80% chance of detecting the specified effect under those assumptions, if that effect and the model assumptions were correct. The 90% power column uses a higher target chance of detection. Neither column describes the probability that the trial’s conclusion is correct. These are prospective planning illustrations calculated from the observed effect, not observed power.

Smaller effects require more participants to detect with the same target power and significance level, because a smaller group difference is harder to distinguish from random variation. For the primary endpoint, planning for 100% of the observed effect corresponds to 272 participants per arm for 80% power and 364 participants per arm for 90% power. Planning for 75% of the observed effect corresponds to 483 and 646 participants per arm, respectively; planning for 50% corresponds to 1,085 and 1,453 participants per arm. The same logic applies to the secondary-outcome planning illustrations, but their assumed effects are specific to each outcome.

These figures should not be read as the trial’s actual design calculation or as evidence about whether the completed trial was adequately powered. They are sensitive to the observed effect used as the starting point, and an observed effect can be an uncertain basis for future planning. The record reports enrollment as 496, but it does not report the original design assumptions that led to that enrollment. A planning illustration and the trial’s actual design rationale are separate pieces of information.

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 = -0.360; pooled standard deviation σ = 1.496.
2. n = 2 × (1.96 + 0.842)² × 1.496² / -0.360² = 271.2, rounded up to 272 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 <- -0.36                      # observed difference in means
sigma <- 1.49645312104264                     # 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(715624)
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 = 272 (80% power), 364 (90% power)
 75% of observed effect: n per arm = 483 (80% power), 646 (90% power)
 50% of observed effect: n per arm = 1,085 (80% power), 1,453 (90% power)
power.t.test: 272.2 
Simulated power at n = 272 per arm: 0.799 (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 analysis is an ANCOVA comparison reported as an LS mean difference of -0.36, with a 95% confidence interval from -0.550 to -0.174 and p < 0.001. The summary-based recalculation estimates a mean difference of -0.36, with an interval from -0.634 to -0.086 and an unadjusted p-value of 0.01. The estimate is the same as rounded, but the uncertainty measures and p-values differ.

The recalculation is unadjusted and uses posted summary statistics; the sponsor’s analysis is model-based. The registry does not provide enough detail to reproduce the ANCOVA, including its covariates and complete model specification. Differences in adjustment and calculation can therefore contribute to different intervals and p-values. The record explicitly flags a discrepancy. In line with that flag, the sponsor’s own analysis takes precedence for interpreting the reported primary result; the recalculation is a supplementary check, not a substitute or correction.

The sponsor result supports a difference in the direction of greater HbA1c reduction in the lixisenatide group under its reported analysis. Its confidence interval communicates uncertainty about the average contrast, while the p-value summarizes evidence relative to the specified null model and analysis assumptions. Neither statistic measures clinical importance or establishes a causal mechanism. Those broader judgments require information beyond the reported test, including the clinical context and the balance of outcomes.

15. Multiplicity, Missing Data and Analysis Populations

Multiplicity arises when several outcomes, time points, subgroups, or statistical comparisons are assessed. Even if each test uses a conventional significance threshold, examining many tests increases the chance that at least one will appear statistically notable by chance. A prespecified testing hierarchy or a multiple-comparison procedure can address this issue, but the registry record does not report whether such a strategy was used for the secondary outcomes. Adjustment for multiplicity is not reported in the registry record.

This matters especially for the secondary glucose comparisons, which are recalculated with unadjusted p-values. Their results should not be read as if each were the only comparison planned. The registry lists additional secondary outcomes, but numerical results for several of them are not reported in the registry record. Without a complete outcome-level analysis plan and results, it is difficult to assess the full scope of testing or whether selective reporting may have affected what is visible.

The primary analysis population is described as mITT: randomized patients who received at least one dose and had baseline and at least one post-baseline efficacy assessment. The primary outcome description also states that LOCF was used. LOCF carries an assumption that the last available measurement is an appropriate substitute for later missing measurements. In general, that assumption can be questionable when outcomes change over time or when missingness is related to the participant’s health or treatment experience. LOCF can affect both estimated changes and uncertainty.

The outcome description defines an on-treatment window and states that observations were to precede rescue therapy. The registry record does not provide a full account of missingness by group, reasons for missing observations, sensitivity analyses, or alternative imputation approaches. Those details are not reported in the registry record. The number analyzed varies by outcome in the posted summaries, but the record does not provide a complete explanation for all outcome-specific analysis populations.

The mITT definition is not identical to analyzing every randomized participant: it requires at least one dose and post-baseline efficacy information. Excluding participants without such assessments may affect comparability if their outcomes differ systematically. The registry’s description enables readers to identify this population rule, but not to assess all its consequences. The statistical analysis plan, adjustment for multiplicity, and comprehensive missing-data sensitivity analyses are not reported in the registry record.

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: coronary artery disease (2.4%), non-cardiac chest pain (1.2%), pneumonia (1.2%), acute myocardial infarction (1.2%), angina pectoris (1.2%).

The registry reports serious adverse events for 17 of 167 participants at risk in the placebo group and 46 of 328 participants at risk in the lixisenatide group. Other adverse events are reported for 104 of 167 participants at risk in the placebo group and 245 of 328 participants at risk in the lixisenatide group. These are event-related counts as recorded in the registry; they should be read with their stated at-risk denominators and not combined into a new total.

The record lists coronary artery disease at 2.4%, and non-cardiac chest pain, pneumonia, acute myocardial infarction, angina pectoris, and coronary angioplasty at 1.2% each among the top serious adverse-event terms. These terms are descriptive listings. They do not establish that the study treatment caused the events, and they do not by themselves indicate comparative risk for a particular event.

Deaths are not reported in the registry record. The record does not provide, in the information available for these counts, the event-by-event timing, severity detail beyond the serious category, or a causal assessment for every event. Safety counts therefore provide a limited descriptive account and should be interpreted separately from the efficacy endpoint. They are not a complete estimate of all possible harms.

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

The reported primary analysis describes an average difference in HbA1c change between randomized groups over the 24-week main treatment period. The sponsor’s ANCOVA estimate is -0.36, with a 95% confidence interval from -0.550 to -0.174. The estimate is in the direction of a greater mean reduction in the lixisenatide group. It does not mean that every participant’s HbA1c decreased, or that individuals experienced the group average.

The result does not establish the size of benefit for a particular person, the probability of an individual response, or whether the difference is important in every clinical context. The registry does not report a threshold for a clinically important HbA1c difference. Statistical evidence, expressed through a confidence interval and p-value, addresses uncertainty under a model; it does not alone answer questions about practical importance, preferences, or overall benefit and harm.

The result also does not settle the findings for every secondary outcome. Some reported glucose measures have estimates in a similar direction, while fasting plasma glucose shows a less distinct difference. Other listed outcomes, including body weight, insulin dose, HbA1c threshold attainment, and rescue therapy, lack numerical results in the registry record. Multiplicity and incomplete reporting limit the strength of broad claims based on secondary findings.

The safety counts describe events recorded in the two groups. They do not prove that treatment caused or prevented an event. Deaths are not reported in the registry record, and the listed counts do not capture every dimension of safety. The study’s extension results are also not reported in the registry record, so the main-period findings should not be treated as a complete account of longer-term outcomes.

Finally, the trial population and treatment context matter. The record concerns people with type 2 diabetes insufficiently controlled with basal insulin, with or without metformin, and a specified add-on regimen. The findings cannot automatically be extended to different populations, background therapies, or treatment schedules. The registry record supports a defined statistical comparison in its stated setting, not a universal conclusion about treatment effects.

18. Why This Trial Matters Statistically

This trial is statistically informative because it sets a randomized, masked comparison around a clearly specified continuous primary endpoint: HbA1c change from baseline to Week 24. The placebo group provides a concurrent comparison for changes that could occur over time or alongside background basal insulin. The sponsor’s ANCOVA estimate and interval give a model-based account of the primary contrast, while the posted group summaries make the direction and scale of change easier to understand.

It also illustrates why analysis details matter. A simple recalculation from summary data produces the same rounded point estimate as the sponsor analysis but a different confidence interval and p-value. The difference is a reminder that a model-adjusted comparison and an unadjusted calculation need not have identical uncertainty. The registry’s discrepancy flag provides a clear reason to give precedence to the sponsor’s reported analysis while retaining the recalculation as a limited check.

The trial also shows how secondary measures can add context without carrying the same evidential weight as a primary endpoint. Glucose measures do not all show the same degree of separation, and unadjusted testing across multiple outcomes raises a multiplicity issue. Missing-data methods and analysis-population rules also shape what the estimates represent; here, the record specifies an mITT population and LOCF but leaves important sensitivity and model details unreported.

Finally, the planning illustrations demonstrate the relationship between effect size and sample size under stated assumptions. They should not be confused with the trial’s original design rationale, which is not reported in the registry record. Together, these features make the study useful for understanding how endpoint choice, adjusted analysis, precision, multiplicity, missing data, and planning assumptions affect interpretation of a clinical trial result.

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 details unavailable. It does not report the original sample-size justification, allocation ratio, full ANCOVA specification, covariates, or model diagnostics. The registry lists quadruple masking but does not describe how masking was maintained or assessed. These omissions limit independent assessment and replication of the primary analysis.

The primary analysis used an mITT population requiring at least one dose and post-baseline efficacy assessment, and LOCF was used. The record does not provide detailed outcome-specific missingness, reasons for missing measurements, or sensitivity analyses using alternative assumptions. The outcome description also places efficacy assessments before rescue therapy, but the consequences of that rule for all participants are not fully reported.

The recalculated primary result is based on summary statistics and approximate variability, rather than participant-level data. It cannot reproduce the sponsor’s adjustment, and the record flags a discrepancy. The sponsor analysis therefore takes precedence. The three numerical secondary results are unadjusted recalculations, and adjustment for multiple testing is not reported in the registry record. Numerical findings for other listed secondary outcomes are also not reported.

Safety information is limited to reported counts and selected serious-event terms. Death counts are not reported in the registry record, and the counts alone do not establish causality. The extension is mentioned in the study description, but its duration and outcomes are not reported in the registry record. These limits constrain conclusions about longer-term effects and the complete balance of benefits and harms.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: GLP-1 Receptor Agonist Lixisenatide in Patients With Type 2 Diabetes for Glycemic Control and Safety Evaluation, on Top of Basal Insulin — NCT00715624.
• Registry record last updated 2016-11-28; status last verified 2016-10.
• Results first posted 2016-10-11.
• 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

The registry record describes a completed, randomized, double-blind, placebo-controlled, parallel-group Phase 3 study of lixisenatide added to basal insulin, with or without metformin. The main treatment period lasted 24 weeks and was followed by an extension. Enrollment was recorded as 496.

The primary endpoint was absolute change in HbA1c at Week 24. The sponsor-reported ANCOVA estimated an LS mean difference of -0.36, with a 95% confidence interval from -0.550 to -0.174 and p < 0.001. A summary-based recalculation differs in interval and p-value; because the discrepancy is flagged, the sponsor analysis takes precedence.

Selected secondary glucose outcomes and adverse-event counts are also reported. Multiplicity adjustment, detailed missing-data sensitivity analyses, the original sample-size justification, and extension results are not reported in the registry record.