Study of Low-Density Lipoprotein Cholesterol (LDL-C) Reduction Using Evolocumab (AMG 145) in Japanese Patients With Advanced Cardiovascular Risk: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Study of Low-Density Lipoprotein Cholesterol (LDL-C) Reduction Using Evolocumab (AMG 145) in Japanese Patients With Advanced Cardiovascular Risk (NCT01953328), a randomized phase 3 study. The primary outcome is Percent Change From Baseline in Low-Density Lipoprotein Cholesterol (LDL-C) at the Mean of Weeks 10 and 12.

2. Clinical Question

The registry describes a 12-week comparison of subcutaneous evolocumab given every 2 weeks or once a month against placebo, in participants receiving statin therapy. The treatment regimens were organized around two atorvastatin doses, 5 mg and 20 mg daily. The treatment period followed a 4-week lipid stabilization period on the assigned atorvastatin dose.

The primary question concerns a continuous laboratory measure rather than a clinical event such as a cardiovascular complication. The comparison is therefore about the difference in average LDL-C change between the randomized regimens over the stated assessment period. It does not directly address longer-term outcomes or effects beyond the treatment period described in the record.

3. Trial Design

This was a phase 3, randomized, triple-masked, parallel-group study sponsored by Amgen. The registry lists enrollment of 409 and a full-analysis-set population for the primary outcome. The study is marked completed. Its listed start was 2013-10, and its primary completion was 2014-06.

Participants received atorvastatin 5 mg or 20 mg daily during a 4-week lipid stabilization period and continued that dose during the 12-week treatment period. Within each atorvastatin dose, the regimens paired either placebo or evolocumab with injections every 2 weeks or once a month. The evolocumab doses were 140 mg every 2 weeks and 420 mg once a month. Placebo injections followed the corresponding schedules.

Randomization and parallel groups mean the trial compared assigned regimens across groups rather than relying on each participant to receive multiple regimens in sequence. Triple masking is the registry's design label; which specific groups of people were masked is not reported in the registry record. Details such as the allocation procedure, stratification factors, and the original sample-size justification are not reported in the registry record.

The primary outcome was measured at baseline and at Weeks 10 and 12, with the endpoint defined as the mean of Weeks 10 and 12. The registry also lists secondary lipid outcomes, including LDL-C, non-HDL-C, apolipoprotein B, and total cholesterol measures. Those laboratory outcomes help describe the pattern across lipid measures, but they remain distinct from clinical outcomes such as cardiovascular events.

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: 409 participants were randomized to 1 of 2 open-label atorvastatin cohorts (5 mg or 20 mg) for the lipid stabilization period; 404 were then randomized to blinded investigational product. Both randomizations were stratified by heterozygous familial hypercholesterolemia (HeFH) diagnosis and lipid-lowering therapy.

Only the first 6 of 8 groups are shown.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Full analysis set

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: Full analysis set (all randomized participants who reveid at least 1 dose of blinded investigational product).

Only the first 6 of 9 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as Percent Change From Baseline in Low-Density Lipoprotein Cholesterol (LDL-C) at the Mean of Weeks 10 and 12, measured over Baseline and Weeks 10 and 12.

The primary endpoint was percent change from baseline in low-density lipoprotein cholesterol (LDL-C) at the mean of Weeks 10 and 12. LDL-C is a continuous laboratory measure, so the outcome captures the average relative change from each participant's baseline rather than whether a participant experienced a specific event. The registry reports group least-squares means and standard errors for this endpoint in the full analysis set.

Using the mean of two follow-up assessments defines the outcome across both Weeks 10 and 12, rather than using only one of those visits. As a general statistical point, averaging measurements can represent the outcome over a specified assessment window, although the exact modeling and handling of measurements are important to interpreting that average.

The endpoint is useful for comparing the randomized regimens on a lipid measure over the stated treatment period. It is not itself a measure of heart attack, stroke, survival, or another clinical event. The registry record does not report a primary clinical cardiovascular-event endpoint, and the short-term lipid endpoint cannot by itself establish effects on such events or on outcomes beyond the follow-up described.

7. Results Posted in the Registry

• A5 Placebo QM: mean 3.91 versus 0.27 on control; difference 3.64 (95% CI -2.32 to 9.60), standardized effect d = 0.24; z-test p = 0.231.
• A5 Evolocumab Q2W: mean -73.70 versus 0.27 on control; difference -73.97 (95% CI -80.17 to -67.77), standardized effect d = -4.70; z-test p = <0.001.
• A5 Evolocumab QM: mean -69.98 versus 0.27 on control; difference -70.25 (95% CI -76.12 to -64.38), standardized effect d = -4.72; z-test p = <0.001.
• A20 Placebo Q2W: mean -0.42 versus 0.27 on control; difference -0.69 (95% CI -8.41 to 7.03), standardized effect d = -0.03; z-test p = 0.861.

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 LDL-C at the Mean of Weeks 10 and 12

• 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:
  • A5 Placebo QM: mean -2.60 versus -14.80 on control; difference 12.20 (95% CI 0.28 to 24.12), standardized effect d = 0.40; z-test p = 0.045.
  • A5 Evolocumab Q2W: mean -104.10 versus -14.80 on control; difference -89.30 (95% CI -101.36 to -77.24), standardized effect d = -2.92; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for A5 Placebo QM 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 A5 Evolocumab Q2W than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Change From Baseline in LDL-C at Week 12

• 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:
  • A5 Placebo QM: mean -0.40 versus -13.70 on control; difference 13.30 (95% CI 0.96 to 25.64), standardized effect d = 0.42; z-test p = 0.035.
  • A5 Evolocumab Q2W: mean -104.50 versus -13.70 on control; difference -90.80 (95% CI -103.55 to -78.05), standardized effect d = -2.81; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for A5 Placebo QM 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 A5 Evolocumab Q2W than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Percent Change From Baseline in Non-HDL-C at the Mean of Weeks 10 and 12

• 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:
  • A5 Placebo QM: mean 4.26 versus 0.18 on control; difference 4.08 (95% CI -1.46 to 9.62), standardized effect d = 0.29; z-test p = 0.149.
  • A5 Evolocumab Q2W: mean -68.48 versus 0.18 on control; difference -68.66 (95% CI -74.63 to -62.69), standardized effect d = -4.53; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between A5 Placebo QM and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for A5 Evolocumab Q2W than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The posted secondary outcomes extend the lipid assessment beyond the primary percent change in LDL-C endpoint. They include absolute change in LDL-C, percent change in non-HDL-C, and measures of apolipoprotein B and total cholesterol, with assessment at the mean of Weeks 10 and 12 or at Week 12. The available secondary results show a general pattern of lower lipid measures in evolocumab groups than in corresponding placebo groups, consistent in direction with the primary endpoint.

These outcomes provide related information about lipid changes; they are not independent confirmation of a clinical cardiovascular benefit. The endpoint measures overlap in what they describe, and the registry's posted results include many outcomes and analyses. The record lists 24 outcome measures and 96 analyses, while the number of secondary outcomes with posted results is 22. The full detail of multiplicity handling is not reported in the registry record.

Unadjusted secondary p-values carry less weight when many outcomes and contrasts are examined. As a general statistical principle, repeated testing increases the opportunity for a nominally small p-value to occur by chance, and correlated endpoints do not remove that concern. Without a reported multiplicity strategy, secondary findings are best treated as supportive or exploratory rather than as a separate set of definitive conclusions. This does not erase their descriptive value; it means the pattern should be considered alongside the prespecified primary result and the analysis plan, details of which are not fully reported.

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

• Change From Baseline in LDL-C at the Mean of Weeks 10 and 12
• Change From Baseline in LDL-C at Week 12
• Percent Change From Baseline in Non-HDL-C at the Mean of Weeks 10 and 12
• Percent Change From Baseline in Non-HDL-C at Week 12
• Percent Change From Baseline in Apolipoprotein B at the Mean of Weeks 10 and 12
• Percent Change From Baseline in Apolipoprotein B at Week 12
• Percent Change From Baseline in Total Cholesterol at the Mean of Weeks 10 and 12
• Percent Change From Baseline in Total Cholesterol at Week 12

9. Statistical Interpretation

10. Statistical Methods Explained

For the primary endpoint, the registry reports least-squares means (LS means), standard errors, and treatment differences from a repeated measures linear effects model. These terms describe model-based summaries of a continuous outcome measured over time. The registry identifies the analysis population as the full analysis set. Further details of the model specification are not reported in the registry record.

What does a least-squares mean represent?

As a general statistical explanation, a least-squares mean is an estimate derived from a fitted model rather than necessarily being the raw arithmetic average of the observed values. A model can account for the structure specified in its analysis plan, such as repeated measurements. The registry record does not identify the covariates or other model terms used, so no additional adjustment should be inferred. The reported LS mean treatment difference is the model's contrast between the two named regimens.

Why use a repeated measures model?

As a general explanation, repeated measures methods are designed for outcomes measured at more than one time point on the same participants. They can use the pattern of measurements across visits to estimate treatment differences over time or at a defined assessment. Here, the primary outcome is defined at the mean of Weeks 10 and 12, and the registry names a repeated measures linear effects model for the sponsor comparisons. The detailed covariance structure, missing-measurement assumptions, and exact model formula are not reported in the registry record.

What does the standard error tell readers?

A standard error describes the estimated sampling uncertainty around a group estimate. It is not the spread of individual participant values. As a general concept, a smaller standard error usually means an estimate is more precise, while a larger standard error indicates more sampling uncertainty. The posted group summaries include standard errors, but they do not provide the raw participant-level data needed to independently reproduce the model analysis.

How should the recalculated comparisons be understood?

The recalculated comparisons use the posted group means and standard errors to form simple, unadjusted contrasts. The pooled standard deviations used for those illustrations are approximate because they are derived from the standard errors, rather than from individual-level variability. This makes them a different analysis from the sponsor's repeated measures model. The recalculated p-values and confidence intervals therefore should not be treated as replacements for the registry's reported model results. Because a discrepancy is flagged, the sponsor's own analysis takes precedence.

As a general statistical point, adjusting for repeated measurements or other prespecified model structure can yield estimates and uncertainty that differ from a simple comparison of group summaries. In addition, rounding of posted estimates and approximations used in the recalculation can contribute to differences. The exact reason for the discrepancy is not reported in the registry record, so it should not be attributed to a particular model feature.

What do p-values and confidence intervals convey?

A p-value describes how compatible the observed data are with a specified null hypothesis under the assumptions of the analysis. It does not measure the probability that the null hypothesis is true, nor does it quantify the clinical importance of an effect. A confidence interval gives a range of values compatible with the data and analysis procedure at its stated confidence level. The registry reports 95% confidence intervals for the sponsor's primary treatment differences.

As a general statistical point, interpreting a set of comparisons also requires attention to multiplicity: the more hypotheses examined, the greater the chance of obtaining small p-values by chance if no multiplicity procedure is used. Whether an adjustment was used for the reported analyses is not reported in the registry record. The number of posted outcomes and analyses is substantial, so the analysis plan and any multiplicity strategy would be useful context.

Several elements that would clarify the analysis are not reported in the registry record: the original sample-size calculation and its assumptions, the handling of missing data, the detailed model specification, and whether interim analyses were planned or conducted. Their absence limits independent assessment of those methodological choices, but does not change the numerical estimates reported for the primary endpoint.

11. Confidence Intervals and Precision

The sponsor-reported primary comparisons include 95% confidence intervals. For the every-2-weeks regimen with atorvastatin 5 mg, the LS mean treatment difference is -73.97 and the interval is -78.54 to -69.41. For the monthly regimen with atorvastatin 5 mg, the estimate is -72.89 and the interval is -77.22 to -68.57. With atorvastatin 20 mg, the every-2-weeks estimate is -74.41, with an interval from -81.21 to -67.61; the monthly estimate is -74.27, with an interval from -78.93 to -69.60. All four reported comparisons have p < 0.001.

A confidence interval expresses uncertainty around an estimate under the statistical model and its assumptions. As a general interpretation, the interval gives a range of effect values that are compatible with the observed data and method; it is not a range containing a stated proportion of individual participant outcomes. The intervals for these primary comparisons are entirely below zero, in keeping with the negative reported treatment differences. Their widths also show that the estimate is not a single exact value.

The posted arm summaries also give standard errors, but those are not interchangeable with the confidence intervals for the treatment contrasts. The latter come from the sponsor's repeated measures analysis. The recalculated intervals based on group summaries are approximate and differ from the sponsor's intervals, so the sponsor-reported intervals take precedence. Confidence intervals do not account for every potential source of uncertainty, including limitations in the design, outcome selection, or analysis assumptions.

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 expressed as participants per arm, not as total trial enrollment or total events. The planning illustration uses a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. It uses the observed effect as its reference and considers effects equal to 100%, 75%, or 50% of that 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 estimated by the planning calculation to give an 80% chance of detecting the stated effect under its assumptions, if that effect is the true effect. The 90% power column uses the same idea but targets a 90% chance. These are planning illustrations, not observed power.

Smaller assumed effects generally require more participants to distinguish them from random variation at the same significance level and power target. For example, the planning entries for 75% or 50% of the observed effect are larger than the entries for the full observed effect in the displayed comparisons. Some values are particularly small or large because these calculations depend on the observed effect and estimated variability; they should not be read as a recommended design without considering a study's scientific aims and assumptions.

The figures are not the enrollment of this completed study and do not account for unreported design choices. They also do not establish that the trial had a particular probability of success after it was conducted. Their purpose is to illustrate how estimated sample requirements change as the target effect and desired power change.

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 = 3.640; pooled standard deviation σ = 15.125.
2. n = 2 × (1.96 + 0.842)² × 15.125² / 3.640² = 271.0, 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 <- 3.64                      # observed difference in means
sigma <- 15.1246532549502                     # 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(1953328)
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), 363 (90% power)
 75% of observed effect: n per arm = 482 (80% power), 646 (90% power)
 50% of observed effect: n per arm = 1,085 (80% power), 1,452 (90% power)
power.t.test: 272 
Simulated power at n = 272 per arm: 0.800 (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 analyses use a repeated measures linear effects model and report LS mean treatment differences, 95% confidence intervals, and p-values. These are the main registry-reported results for comparing evolocumab with placebo within the matching atorvastatin-dose and injection-schedule regimens.

The recalculated comparisons are simple, unadjusted estimates derived from the posted arm summaries. They rely on approximate pooled variability inferred from standard errors and do not reproduce the sponsor's repeated measures model. The registry flags a discrepancy between the recalculations and sponsor analyses. Accordingly, the sponsor's own analysis takes precedence; the recalculations are illustrative and should not replace the reported model estimates or their confidence intervals.

This distinction matters because a treatment contrast and its uncertainty depend on the analysis method, not only on the group means. The sponsor estimates are the appropriate values to quote for the primary inferential results. The recalculations can help readers see the basic direction and scale of a contrast, but they are not an independent replication of the trial analysis. The reason for the discrepancy is not reported in the registry record, so it cannot be assigned to a specific modeling decision.

15. Multiplicity, Missing Data and Analysis Populations

The registry lists 24 outcome measures and 96 analyses, and it reports 22 secondary outcomes with posted results. Such a broad set of outcomes and comparisons raises a multiplicity question: when many statistical hypotheses are tested, the chance of at least one apparently positive result can increase. As a general statistical explanation, a prespecified adjustment or hierarchical testing plan can help control that risk, but the method must be known to interpret the collection of p-values.

Whether the trial used a multiplicity adjustment, a testing hierarchy, or another strategy is not reported in the registry record. The reported primary comparisons have p < 0.001, but the record does not give enough information to say whether or how those values were adjusted for the set of analyses. The secondary results should therefore be interpreted with particular caution as a group, rather than treating each nominal p-value as an isolated confirmatory finding.

The primary endpoint population is identified as the full analysis set. The registry record does not report enough detail to determine the precise inclusion rules for that set, how missing endpoint measurements were handled, or whether any sensitivity analyses assessed alternative missing-data assumptions. Missing data can affect estimates if measurement availability is related to treatment or outcome; this is a general methodological concern, not a claim that such bias occurred in this trial.

The detailed statistical analysis plan, model specification, and any interim-analysis procedures are not reported in the registry record. These omissions limit the ability to assess prespecification, analysis choices, and the handling of multiple comparisons independently. They do not justify assuming that a particular procedure was or was not used.

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: epilepsy (2.0%), pneumonia bacterial (2.0%), contusion (2.0%), endometrial cancer (2.0%), rotator cuff syndrome (2.0%).

The registry lists serious adverse events by treatment group and reports the number of participants at risk and affected. The serious-event counts are 1 in the A5 placebo every-2-weeks group, 2 in the A5 placebo monthly group, 0 in the A5 evolocumab every-2-weeks group, and 1 in the A5 evolocumab monthly group. In the A20 groups, the serious-event counts are 1 for placebo every 2 weeks, 1 for placebo monthly, 0 for evolocumab every 2 weeks, and 0 for evolocumab monthly. These are descriptive counts, not a formal comparison of risk.

The registry also lists other adverse events by group, with affected counts ranging from 6 to 14 across the listed regimens. The listed serious-event terms include epilepsy, bacterial pneumonia, contusion, endometrial cancer, rotator cuff syndrome, and brain stem infarction, each with a percentage reported in the registry. These entries do not establish that a treatment caused an event.

Death counts are not reported in the registry record. The record's adverse-event information should be read alongside the duration of the reported treatment and follow-up; it does not, by itself, establish comparative safety over longer periods.

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

The results describe short-term changes in lipid measures among Japanese adults with high cardiovascular risk and hyperlipidemia or mixed dyslipidemia who received atorvastatin. The registry-reported primary comparisons show negative LS mean differences for each evolocumab regimen against its matched placebo regimen, with all four reported p-values below 0.001 and 95% confidence intervals that do not include zero. The estimates quantify differences in percent change in LDL-C at the mean of Weeks 10 and 12.

These findings are specific to the regimens, population, and time period studied. They do not establish how the regimens compare with one another, because the reported primary analyses are each against a corresponding placebo group. They also do not establish the durability of lipid changes beyond the reported assessment period.

The endpoint is a laboratory measure, not a clinical event. The results do not by themselves show whether cardiovascular events, survival, symptoms, or quality of life differ. The registry record does not report a primary cardiovascular-event endpoint. The findings also should not be generalized automatically to populations or treatment settings that were not described in the trial record.

Statistical significance is not the same as clinical importance. Confidence intervals characterize uncertainty under the reported model, while broader interpretation also depends on study conduct, missing-data handling, multiplicity, and the relevance of the outcome to clinical decisions. Some of those methodological details are not reported in the registry record.

18. Why This Trial Matters Statistically

This trial offers a clear example of why a treatment contrast should be interpreted in the context of its design and endpoint. It randomized participants to parallel regimens, paired evolocumab or placebo with atorvastatin, and assessed a continuous lipid outcome at a defined period. The primary analysis is reported as a repeated measures model, which is more informative for inference than simply comparing the posted arm means without regard to the analysis method.

It also illustrates the difference between a within-group change and a between-group treatment contrast. A group mean describes its change from baseline; the randomized comparison asks how that change differs from the corresponding placebo regimen. The reported LS mean treatment differences and confidence intervals directly address that comparative question.

The discrepancy between sponsor analyses and simple recalculations is a useful statistical caution: the same posted summaries do not necessarily reproduce an analysis based on a fuller model. The registry's sponsor-reported analysis therefore has priority over the illustrative recalculation. In addition, the number of listed analyses makes multiplicity and prespecification relevant to the interpretation of secondary results.

Finally, the trial separates biomarker evidence from evidence about clinical events. The results characterize lipid changes over the stated period; they do not answer every question about longer-term outcomes or safety. Recognizing the endpoint's scope is essential to drawing conclusions that match what the design measured.

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 treatment period and primary assessment are limited to 12 weeks, with the primary endpoint assessed at the mean of Weeks 10 and 12. The results therefore characterize short-term lipid changes, not long-term persistence. The registry's listed outcomes focus on lipid measures and do not include a reported primary cardiovascular-event endpoint.

The study population is described as Japanese adults with high cardiovascular risk and hyperlipidemia or mixed dyslipidemia receiving atorvastatin. Generalization beyond that population and treatment context is not established by the record. The trial compares particular evolocumab schedules with matching placebo schedules; the reported primary comparisons do not directly test every schedule against every other schedule.

Several details needed for a fuller appraisal are not reported in the registry record: the original sample-size justification, model specification beyond the repeated measures linear effects model label, missing-data methods, interim analyses, and multiplicity strategy. The full-analysis-set label is reported, but its detailed inclusion rules are not. These omissions limit independent assessment of the analysis plan and sensitivity to assumptions.

The simple recalculations are approximate and differ from sponsor-reported model analyses. The sponsor analyses take precedence, but the reason for the discrepancy is not reported. Adverse-event counts are descriptive, and deaths are not reported in the registry record. Taken together, these limits argue for keeping conclusions tied to the posted endpoint, analysis, population, and follow-up rather than extending them beyond what was measured.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Study of Low-Density Lipoprotein Cholesterol (LDL-C) Reduction Using Evolocumab (AMG 145) in Japanese Patients With Advanced Cardiovascular Risk — NCT01953328.
• Registry record last updated 2015-12-23; status last verified 2015-11.
• Results first posted 2015-12-23.
• 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

NCT01953328 is a completed, randomized, triple-masked, parallel-group phase 3 study sponsored by Amgen. The registry lists enrollment of 409 and a study population of Japanese adults with high cardiovascular risk and hyperlipidemia or mixed dyslipidemia.

Participants received atorvastatin 5 mg or 20 mg daily and were assigned to placebo or evolocumab injections every 2 weeks or once a month for 12 weeks, after a 4-week lipid stabilization period. The primary endpoint was percent change from baseline in LDL-C at the mean of Weeks 10 and 12, analyzed in the full analysis set.

The sponsor-reported repeated measures analyses show negative treatment differences for each evolocumab regimen compared with its matching placebo regimen; all have p < 0.001. These are short-term lipid results. The record does not report a primary clinical cardiovascular-event endpoint, the original sample-size justification, or several details of the statistical plan.