Alzheimer Disease Proof of Concept Study With BI 409306 Versus Placebo: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Alzheimer Disease Proof of Concept Study With BI 409306 Versus Placebo (NCT02240693), a randomized phase 2 study. The primary outcome is Change From Baseline in Neuropsychological Test Battery in Total Z-score After 12-week Treatment..

2. Clinical Question

The record describes a phase 2 proof-of-concept study intended to compare the effects of four oral BI 409306 doses with placebo. Its primary endpoint was a continuous measure, reported as a least-squares mean with a standard error for each group. The registry record reports separate sponsor analyses for the dose comparisons and for pooled BI 409306. Those analyses estimate between-group differences, with confidence intervals and p-values.

The trial also assessed activities of daily living, dementia-rating scale scores, and a cognitive scale as secondary outcomes. Together, these measures provide additional perspectives on change during treatment, but they do not replace the registered primary endpoint. The registry record does not report a broader clinical question, such as long-term outcomes, beyond the 12-week treatment comparison.

3. Trial Design

This was a randomized, double-blind, parallel-group phase 2 study sponsored by Boehringer Ingelheim. The record lists Alzheimer disease as the condition and oral BI 409306 and matching placebo as the interventions. The study was designed to compare four BI 409306 dose regimens with placebo during a 12-week treatment period. Its enrollment was 128, and its status is recorded as completed.

The listed regimens were BI 409306 10 mg once daily, 25 mg once daily, 50 mg once daily, and 25 mg twice daily. Results are reported for each regimen and for pooled BI 409306. Placebo results are reported as the comparator. Randomization and double masking are stated in the record; further details about sequence generation, concealment, or how masking was maintained are not reported in the registry record.

The study began on 2015-01-15, with primary completion on 2017-09-18. The primary efficacy population was the full analysis set: randomized patients who received at least one dose and had a baseline and at least one post-baseline on-treatment efficacy assessment. The primary endpoint used observed cases. More detailed rules for handling missing measurements, discontinuation, or other protocol deviations are not reported in the registry record.

The record lists three secondary outcomes: change in ADCS-MCI-ADL total score, Clinical Dementia Rating Scale Sum of Boxes total score, and ADAS-cog11 total score after 12-week treatment. It reports five outcome measures and 22 analyses overall. The original sample-size justification and detailed statistical analysis plan 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: For this trial, patients were randomised at 36 sites in 11 countries. Following an initial Screening Visit and a single blinded 2-week placebo run-in period, patients who qualified according to the in- and exclusion criteria were randomised to one of the five treatment groups

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The full analysis set (FAS) included all randomised patients who were treated with at least one dose of trial medication and had a baseline and at least one post-baseline on-treatment efficacy assessment. Observed cases (OC)

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

Population: Treated set: The treated set (TS) included all patients who were randomised and treated with at least one dose of trial medication.

6. Primary Endpoint

The registry lists the primary outcome as Change From Baseline in Neuropsychological Test Battery in Total Z-score After 12-week Treatment., measured over Baseline and 12 weeks.

The primary endpoint was change from baseline in Neuropsychological Test Battery total z-score after 12 weeks of treatment. A z-score expresses a value in relation to a reference mean in standard-deviation units. The record states that a score of 0 is equal to the mean at baseline; negative values indicate values lower than baseline, while positive values indicate values higher than baseline. Change was calculated as the post-baseline composite z-score minus the pretreatment z-score, so a positive change indicates improvement from baseline.

The endpoint is continuous, meaning it can take a range of values rather than simply classifying each participant into an event category. A group mean summarizes the average change, while its standard error describes the estimated uncertainty around that mean. The sponsor reported least-squares means and used a mixed model repeated-measures analysis for the primary comparisons.

A standardized score can make a composite measure interpretable relative to its baseline distribution. It does not, by itself, establish how meaningful a particular change is to patients. The registry record does not report a threshold for a clinically important difference, nor does it provide a detailed account of the individual tests making up the composite. The endpoint therefore supports a statistical comparison of average change during the treatment period, not a stand-alone conclusion about broader or longer-term outcomes.

7. Results Posted in the Registry

• BI 409306 10 Milligram (mg) Once Daily (QD): mean 0.35 versus 0.27 on control; difference 0.08 (95% CI -0.07 to 0.23), standardized effect d = 0.30; z-test p = 0.284.
• BI 409306 25 mg QD: mean 0.20 versus 0.27 on control; difference -0.07 (95% CI -0.22 to 0.08), standardized effect d = -0.26; z-test p = 0.359.
• BI 409306 50 mg QD: mean 0.26 versus 0.27 on control; difference -0.01 (95% CI -0.16 to 0.14), standardized effect d = -0.04; z-test p = 0.898.
• BI 409306 25 mg Twice Daily (BID): mean 0.32 versus 0.27 on control; difference 0.05 (95% CI -0.10 to 0.20), standardized effect d = 0.19; z-test p = 0.517.

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 ADCS-MCI-ADL (Alzheimer's Disease Cooperative Study/Activities of Daily Living for Patients With Mild Cognitive Impairment) Total Score After 12-week Treatment

• 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:
  • BI 409306 10 Milligram (mg) Once Daily (QD): mean 0.24 versus 0.38 on control; difference -0.14 (95% CI -2.30 to 2.02), standardized effect d = -0.04; z-test p = 0.899.
  • BI 409306 25 mg QD: mean 1.79 versus 0.38 on control; difference 1.41 (95% CI -0.79 to 3.61), standardized effect d = 0.36; z-test p = 0.209.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between BI 409306 10 Milligram (mg) Once Daily (QD) and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between BI 409306 25 mg QD and the control. The interval gives the range of effects that remain compatible with the data.

Change From Baseline in Clinical Dementia Rating Scale Sum of Boxes (CDR-SB) Total Score After 12-week Treatment

• 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:
  • BI 409306 10 Milligram (mg) Once Daily (QD): mean 0.00 versus 0.10 on control; difference -0.10 (95% CI -0.56 to 0.36), standardized effect d = -0.11; z-test p = 0.672.
  • BI 409306 25 mg QD: mean 0.40 versus 0.10 on control; difference 0.30 (95% CI -0.18 to 0.78), standardized effect d = 0.34; z-test p = 0.219.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between BI 409306 10 Milligram (mg) Once Daily (QD) and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between BI 409306 25 mg QD and the control. The interval gives the range of effects that remain compatible with the data.

Change From Baseline in Alzheimer's Disease Assessment Scale-cognitive Subscale (ADAS-cog11) Total Score After 12-week Treatment

• 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:
  • BI 409306 10 Milligram (mg) Once Daily (QD): mean 0.98 versus 1.24 on control; difference -0.26 (95% CI -2.38 to 1.86), standardized effect d = -0.07; z-test p = 0.810.
  • BI 409306 25 mg QD: mean 0.62 versus 1.24 on control; difference -0.62 (95% CI -2.82 to 1.58), standardized effect d = -0.16; z-test p = 0.580.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between BI 409306 10 Milligram (mg) Once Daily (QD) and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between BI 409306 25 mg QD and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcomes cover activities of daily living, dementia-rating scale scores, and a cognitive scale. Across these measures, the posted group means and comparisons do not form a consistent pattern that clearly reinforces the primary result. Estimates vary by dose and outcome, and the sponsor-reported comparisons available for these outcomes have confidence intervals that include zero. The registry record posts results for three secondary outcomes, but not every possible comparison is represented in the reported analyses.

These measures provide complementary descriptions of change over the treatment period. They do not independently establish the primary endpoint result, and they should not be treated as substitutes for it. The primary endpoint was the registered main outcome; the secondary measures add context but have less inferential weight when multiple comparisons are considered.

When several doses and outcomes are examined, some estimates may appear notable by chance alone, even if there is no consistent underlying difference. Unless a testing hierarchy or adjustment is specified, unadjusted p-values do not account for the full set of comparisons. The registry record does not report a multiplicity procedure for the secondary outcomes. Accordingly, their results are best interpreted as supportive or exploratory patterns, not as definitive evidence for a separate conclusion.

In general, coherence across outcomes can be informative, but it must be judged alongside the uncertainty of each estimate, the outcome's measurement properties, and the analysis plan. The posted secondary results do not resolve the uncertainty in the primary comparison. Detailed protocol-defined roles for the secondary endpoints, including whether any were intended as confirmatory, are not reported in the registry record.

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

• Change From Baseline in ADCS-MCI-ADL (Alzheimer's Disease Cooperative Study/Activities of Daily Living for Patients With Mild Cognitive Impairment) Total Score After 12-week Treatment
• Change From Baseline in Clinical Dementia Rating Scale Sum of Boxes (CDR-SB) Total Score After 12-week Treatment
• Change From Baseline in Alzheimer's Disease Assessment Scale-cognitive Subscale (ADAS-cog11) Total Score After 12-week Treatment

9. Statistical Interpretation

10. Statistical Methods Explained

The posted results combine a sponsor-reported analysis with recalculations based on group-level summary statistics. They answer related questions, but their methods differ. The distinction matters: the sponsor analysis is the trial's reported inferential analysis, whereas a recalculation from means and standard errors cannot reproduce every feature of a participant-level model.

What does a mean difference represent?

A mean difference compares the average outcome in one group with the average in another. For the primary endpoint, the sponsor reported differences in final values between each BI 409306 regimen and placebo, as well as between pooled BI 409306 and placebo. A positive estimate indicates a higher value in the first-listed treatment group relative to placebo; a negative estimate indicates a lower value. The endpoint's definition says positive change from baseline indicates improvement, but that direction should not be confused with a finding of a between-group difference.

The recalculated estimates use the posted group means and standard errors. They are unadjusted approximations and do not incorporate the same participant-level repeated-measures modeling as the sponsor's analysis. They can be useful for understanding the scale and uncertainty of a contrast, but should not be treated as a replacement for the reported analysis.

Why use a mixed model repeated-measures analysis?

A mixed model repeated-measures analysis, or MMRM, is a modeling approach for outcomes measured over repeated time points. It can model the pattern of measurements over time and account for correlations among measurements from the same participant. Depending on its specification, it can use available repeated observations without reducing the data to a single endpoint measurement for every participant. The precise model specification, including covariance structure and missing-data assumptions, is not reported in the registry record.

The sponsor used MMRM for the primary endpoint. This is different from the recalculation, which works from the posted summary statistics rather than the original sequence of participant measurements. It is therefore expected that the estimates, confidence limits, and p-values need not match exactly. The record does not indicate a discrepancy flag, but the trial-reported MMRM results remain the appropriate primary results to interpret.

What do standard errors and confidence intervals tell us?

A standard error describes uncertainty in an estimated mean. It is not the spread of individual participant outcomes. A confidence interval combines an estimate and its uncertainty to show a range of values that are compatible with the data under the analysis assumptions. Wider intervals indicate less precision; narrower intervals indicate greater precision. For the primary between-group comparisons, the posted 95% confidence intervals all include zero, leaving uncertainty about whether the difference favors either group or is close to zero.

The confidence interval is not a guarantee that the true value lies within the particular interval, nor does it measure the probability that a hypothesis is true. Its interpretation depends on the statistical model and its assumptions. In this trial, the sponsor's intervals come from MMRM. The recalculated intervals are approximate and based on posted summary statistics.

What does a p-value contribute?

A p-value describes how unusual data at least as incompatible with a specified null hypothesis would be, assuming that hypothesis and the model assumptions. It is not the probability that the treatment has no effect, nor a measure of clinical importance. The primary sponsor-reported p-values are all greater than 0.05. Read alongside the estimates and intervals, they indicate that the results do not sharply distinguish the observed comparisons from a zero difference under the reported analysis.

The record labels the analyses as superiority comparisons. A failure to demonstrate superiority is not a demonstration that groups are equivalent. Equivalence or non-inferiority requires a prespecified margin and an analysis designed to address that question; such a margin is not reported in the registry record.

How should pooled results be understood?

The pooled BI 409306 result combines the active-treatment arms for a comparison with placebo. Pooling can yield a more precise estimate than an individual dose comparison, but it also answers a broader question: the average comparison across the pooled regimens. It does not show that every dose has the same effect or that the dose-specific results are interchangeable. The criteria for pooling and any dose-response modeling are not reported in the registry record.

Finally, no adjustment for multiple comparisons is reported. Several dose groups and outcomes were examined, so chance findings become more likely when many hypotheses are tested. This is particularly relevant to secondary outcomes and exploratory comparisons. The registry record does not report a prespecified hierarchy or a multiplicity-control procedure, and the presented p-values should be interpreted with that uncertainty in mind.

11. Confidence Intervals and Precision

The sponsor-reported primary comparisons have 95% confidence intervals that include zero. For BI 409306 10 mg once daily, the reported mean difference was 0.07, with an interval from -0.074 to 0.223. For 25 mg once daily, the estimate was -0.07, with an interval from -0.220 to 0.082. For 50 mg once daily, it was -0.02, with an interval from -0.171 to 0.139. For 25 mg twice daily, it was 0.05, with an interval from -0.106 to 0.199. The pooled estimate was 0.01, with an interval from -0.092 to 0.121.

These intervals are compatible with differences in either direction and with values near zero. Their span is important: the estimates alone may look close to zero, but an estimate is only one point within a broader range of uncertainty. The results therefore do not pin down a narrow range for the between-group difference. The record does not provide a clinically important difference threshold, so it is not possible to judge these intervals against a registry-stated threshold for meaningful change.

The recalculated intervals are also wide enough to include zero. They are based on approximate, unadjusted calculations from the posted group summaries and should be distinguished from the sponsor's MMRM intervals. The sponsor analysis takes precedence for interpretation of the trial's primary comparison.

In general, precision depends on the amount and variability of information available and on the analysis. A confidence interval is not a statement that there is no difference; it describes the range of differences that remain compatible with the data and model. The reported results support uncertainty about the size and direction of any primary between-group difference rather than a precise estimate of a treatment effect.

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 sample-size illustrations express the number of participants per arm for a future two-group comparison. They are planning illustrations based on the observed effect in this trial, not the original design calculation. The listed assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. For each comparison, the illustration considers detecting the observed effect or a stated fraction of that effect. Smaller effects need more participants to detect under the same assumptions.

The 80% power column gives the estimated participants per arm for a design with an 80% chance of detecting the specified effect at the stated significance level, if that effect and the assumptions are correct. The 90% power column gives the corresponding planning figure for a 90% chance. Greater planned power generally requires more participants. Each row describes planning for 100%, 75%, or 50% of the observed effect; the smaller-effect rows require larger per-arm numbers.

These figures are not observed power. They do not measure the chance that this completed trial found an effect, and they should not be used to reinterpret its p-values. Because the observed estimates are uncertain, using them as planning targets is also uncertain. The original sample-size justification and assumptions for the trial itself are not reported in the registry record.

The registry record describes continuous outcomes, so the displayed planning figures refer to participants per arm, not total events. The calculations are illustrations rather than recommendations. A future study would need a justified target effect and assumptions appropriate to its own design, population, outcome, and analysis.

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 = 0.080; pooled standard deviation σ = 0.270.
2. n = 2 × (1.96 + 0.842)² × 0.270² / 0.080² = 178.8, rounded up to 179 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 <- 0.08                      # observed difference in means
sigma <- 0.269964195156815                     # 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(2240693)
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 = 179 (80% power), 240 (90% power)
 75% of observed effect: n per arm = 318 (80% power), 426 (90% power)
 50% of observed effect: n per arm = 716 (80% power), 958 (90% power)
power.t.test: 179.7 
Simulated power at n = 179 per arm: 0.798 (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 used mixed model repeated-measures methods and reported mean differences in final values with 95% confidence intervals and p-values. Its estimates range from -0.07 to 0.07 for the dose-specific comparisons, while the pooled comparison estimate is 0.01. Each reported confidence interval includes zero.

The recalculated comparisons use the posted group means and standard errors to make simple, unadjusted comparisons. For the same primary endpoint, these recalculated estimates are 0.08 for 10 mg once daily, -0.07 for 25 mg once daily, -0.01 for 50 mg once daily, 0.05 for 25 mg twice daily, and 0.02 for pooled BI 409306. These values are close to the sponsor estimates, but they are not identical, and their p-values and intervals should not be substituted for the sponsor's reported results.

The distinction arises because the recalculation does not reproduce the sponsor's repeated-measures model or use the participant-level data. It is a summary-statistic approximation, not a second trial analysis. The sponsor's own analysis takes precedence for interpreting the trial results. Both sets of comparisons indicate uncertainty around the estimates, but conclusions should be anchored in the reported MMRM analysis and its confidence intervals.

15. Multiplicity, Missing Data and Analysis Populations

The record lists four dose regimens, a pooled active-treatment comparison, and multiple outcomes. Each additional comparison creates another opportunity for a result to appear unusual by chance. Statistical methods can address this multiplicity through a prespecified testing hierarchy or an adjustment to the decision thresholds, but the registry record does not report such a procedure for the results.

The reported primary p-values should therefore be interpreted as the p-values for the individual reported comparisons, not as evidence that a family of tests was controlled for multiplicity. The registry record does not report whether dose comparisons were ordered, whether a single primary comparison was given priority, or whether any other procedure was used. This uncertainty is especially relevant when interpreting secondary outcomes and exploratory comparisons. A collection of unadjusted results is weaker evidence than a single prespecified test with a clearly described analysis plan.

The primary analysis population is described as the full analysis set: randomized patients who received at least one dose and had a baseline and at least one post-baseline on-treatment efficacy assessment. The primary outcome used observed cases. This definition clarifies which patients were eligible for the efficacy analysis, but the record does not report the detailed handling of missing measurements, reasons for missingness, sensitivity analyses, or assumptions used by the MMRM model. The exact analysis population and available observations can affect precision and interpretation.

The statistical analysis plan, the trial's original sample-size justification, and any interim analyses are not reported in the registry record. No assumptions should be inferred about their use. Without those details, readers cannot fully assess whether all reported analyses were prespecified or how the sponsor handled alternative analytic choices. These gaps do not invalidate the posted estimates, but they limit assessment of the complete inferential framework.

16. Safety Results

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

Deaths

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: haematuria (4.8%), abdominal pain (2.3%).

The registry record lists serious adverse events for each treatment group. No serious adverse events were reported among participants at risk in the 10 mg once-daily, 25 mg once-daily, or 50 mg once-daily groups. One was reported in the 25 mg twice-daily group, and one in the matching-placebo group. No deaths were reported in any listed group.

The record also gives serious-event terms with percentages: Haematuria at 4.8% and Abdominal pain at 2.3%. These are the reported terms and percentages; the registry record does not provide further detail about their timing, severity, relationship to treatment, or how the percentages were derived in the displayed summary.

Serious adverse events are not the same as all adverse events. For other adverse events, the record reports 5 affected in the 10 mg once-daily group, 8 in the 25 mg once-daily group, 11 in the 50 mg once-daily group, 3 in the 25 mg twice-daily group, and 6 in the placebo group. These are descriptive counts, not adjusted comparisons. Differences in group sizes and the absence of detailed event-level information limit direct comparison. The registry record does not report enough context to attribute events to treatment or to draw a broader safety conclusion.

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

The primary results estimate average differences in Neuropsychological Test Battery total z-score after 12 weeks. The sponsor-reported estimates are close to zero, and all the reported 95% confidence intervals include zero. The data therefore do not make a clear distinction between the compared groups on this endpoint, while still allowing a range of possible differences in either direction.

This is not the same as demonstrating that the groups are equivalent. A conclusion of equivalence requires a prespecified margin defining what differences would be considered sufficiently small, together with an analysis designed to test that question. The registry record does not report such a margin. Nor does a non-significant superiority comparison prove that no effect exists; it indicates that the reported analysis did not provide strong evidence against the no-difference hypothesis at the stated testing framework.

The endpoint's direction is defined in the record: positive change indicates improvement from baseline. That definition helps interpret the scale, but statistical differences alone do not establish whether a change is clinically important. A threshold for meaningful change is not reported in the registry record. The estimates also relate to the stated analysis population, observed cases, and 12-week treatment period.

The secondary outcomes add context but do not remove uncertainty in the primary result, especially because the record does not report a multiplicity adjustment or a confirmatory hierarchy. The safety summary is descriptive and does not establish causation. The posted data should therefore be understood as a set of trial estimates with uncertainty, not as medical advice or a conclusion about treatment for an individual. Longer-term effects and outcomes beyond those reported are not addressed by these results.

18. Why This Trial Matters Statistically

This trial illustrates why a treatment-group mean and a between-group comparison answer different questions. Each arm's mean describes change from its own baseline, while the comparison estimates how the groups differ. The primary endpoint group means are positive in several arms, but those within-group changes alone cannot show whether BI 409306 differs from placebo. The estimated contrasts and their confidence intervals are needed for that interpretation.

It also shows the value of reading an estimate together with its uncertainty. The sponsor-reported primary estimates are small in magnitude, but their confidence intervals span values on both sides of zero. A p-value by itself would not convey the range of differences compatible with the data. The interval makes the limits of precision visible and guards against treating a point estimate as an exact answer.

The comparison between sponsor analyses and summary-statistic recalculations is another useful statistical lesson. A simple recalculation can help readers understand the scale of a contrast, but it cannot reproduce a repeated-measures model using participant-level data. Small differences between estimates do not necessarily signal a contradiction; they can reflect different methods. For interpretation, the trial-reported analysis takes precedence.

Finally, the study has multiple doses and outcomes. That structure creates questions about multiplicity, prespecification, and the role of secondary endpoints. The registry record does not report those details, so the results require cautious interpretation. These features make the trial a useful example of separating descriptive summaries, primary inference, exploratory evidence, and planning illustrations.

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 does not report the original sample-size justification, the full statistical analysis plan, or detailed model specifications for the sponsor's MMRM. It also does not describe interim analyses, multiplicity control, or a testing hierarchy. These omissions limit the ability to assess how the primary and secondary analyses were planned and how all comparisons should be interpreted together.

The primary analysis used observed cases in a defined full analysis set, requiring a baseline and at least one post-baseline on-treatment efficacy assessment. The detailed missing-data rules, reasons for missingness, and sensitivity analyses are not reported in the registry record. As a result, the extent to which missing measurements could affect estimates cannot be evaluated from the reported information alone.

The recalculated comparisons are approximate and unadjusted, based on posted summary statistics. They do not reproduce the sponsor's participant-level repeated-measures analysis. Their purpose is to provide a simple comparison, not to replace the MMRM results. The sponsor analysis should be used for the primary interpretation.

The registry record reports a 12-week treatment period and does not establish longer-term outcomes. It also does not report a clinically important threshold for the primary measure, so the practical meaning of the estimated differences cannot be assessed against a stated benchmark. The listed safety results are descriptive and lack detail on event timing, attribution, and clinical context. These limitations apply to interpretation of the posted record and should not be filled by assumptions about the protocol or analyses.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Alzheimer Disease Proof of Concept Study With BI 409306 Versus Placebo — NCT02240693.
• Registry record last updated 2018-11-28; status last verified 2018-10.
• Results first posted 2018-11-28.
• 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

The registry record describes a completed, randomized, double-blind, parallel-group phase 2 study of oral BI 409306 versus matching placebo in patients with Alzheimer disease. Four BI 409306 regimens were assessed during a 12-week treatment period, with results also reported for pooled BI 409306.

The primary outcome was change from baseline in Neuropsychological Test Battery total z-score. Sponsor-reported MMRM comparisons had 95% confidence intervals that included zero. The posted record also reports secondary outcomes covering activities of daily living, dementia-rating scale scores, and a cognitive scale, along with descriptive adverse-event information.

The primary analysis population and observed-case approach are described, but the original sample-size rationale, detailed missing-data methods, and multiplicity procedure are not reported in the registry record. Summary-statistic recalculations are approximate and do not replace the sponsor's analysis.