Efficacy and Safety of LCZ696 Compared to Valsartan on Cognitive Function in Patients With Chronic Heart Failure and Preserved Ejection Fraction: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Efficacy and Safety of LCZ696 Compared to Valsartan on Cognitive Function in Patients With Chronic Heart Failure and Preserved Ejection Fraction (NCT02884206), a randomized phase 3 study. The primary outcome is Change From Baseline in the CogState Global Cognitive Composite Score (GCCS).

2. Clinical Question

3. Trial Design

This was a phase 3, randomized, double-blind, parallel-group, active-controlled study sponsored by Novartis Pharmaceuticals. The registry records an enrollment of 592. The study began on 2016-11-23, reached primary completion on 2022-05-16, and is listed as completed. The primary endpoint timeframe was baseline and month 36.

The treatment sequence included a single-blind run-in. The arm descriptions state that patients who tolerated treatment during the run-in were randomized afterward to LCZ696 200 mg twice daily or valsartan 160 mg twice daily, with treatment planned for three years. This sequence matters when interpreting who entered the randomized comparison: the arm descriptions refer to run-in tolerance, while detailed criteria for tolerability and for study eligibility are not reported in the registry record.

The primary outcome analysis included patients in the Safety Set with a non-missing CogState GCCS value at both baseline and month 36. The Safety Set is defined in the record as randomized patients who received at least one dose of study drug. The primary arm summaries list 195 participants for LCZ696 and 183 for valsartan. Other analyses can have different available-participant counts; for example, the posted IADL summaries list 197 and 186, respectively. These are outcome-specific analysis counts, not a basis for inferring why data were missing or how many participants completed all assessments.

The record describes four outcome measures and six analyses, with three secondary outcomes posted. It does not report the original design assumptions, the planned sample-size justification, or whether interim analyses were conducted. Those details should not be inferred from the enrollment or from the planning illustrations based on the observed effect.

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 1200 patients were screened. Of them, 706 patients entered the run-in period. Patients entered different parts of the treatment run-in phase based on their use of renin angiotensin system blockade medications at the time of enrolment (valsartan 40 mg bid (if necessary), followed by valsartan 80 mg bid, and then followed by LCZ696 100 mg bid, over 3 to 8 weeks duration)

Valsartan Run-in Period

Reasons for not completing this period:

LCZ696 Run-in Period

Reasons for not completing this period:

Double-Blind Period

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The overall number of participants analyzed represents the participants in the Safety Set (SAF) with non-missing value of CogState GCCS at both, baseline and month 36. The SAF comprised all randomized patients who received at least one dose of study drug.

5. Baseline Characteristics

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

Population: Baseline Participants are based on the Randomized population, as primary and secondary analysis are based on the Randomized Set

6. Primary Endpoint

The registry lists the primary outcome as Change From Baseline in the CogState Global Cognitive Composite Score (GCCS), measured over Baseline, month 36.

The primary endpoint was change from baseline in the CogState Global Cognitive Composite Score (GCCS), assessed at baseline and month 36. CogState comprised seven tests administered electronically at scheduled visits. For each test, a standardized z-score was calculated, and the GCCS was the average of the non-missing individual test z-scores. A higher score indicates better cognitive function.

A composite score can provide a single summary across several cognitive tasks rather than treating each task as the sole primary measure. Standardizing the component test results allows them to contribute to a common summary, while averaging non-missing test scores means that the composite may be based on fewer than all component tests for a participant. The registry record does not give further details on how many component scores were required for a valid composite.

The endpoint is a continuous measure, and the posted arm values are least-squares mean changes with standard errors. The registry describes a repeated-measures ANCOVA for the primary analysis, using measurements across visits and adjusting for baseline GCCS and specified factors. The endpoint addresses measured cognitive test performance over the study period; it is not itself a direct measure of every aspect of daily function or clinical status.

7. Results Posted in the Registry

• Valsartan 160 mg Bid: mean -0.07 versus -0.09 on control; difference 0.02 (95% CI -0.09 to 0.12), standardized effect d = 0.03; z-test p = 0.736.

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 the Summary Score of the Instrumental Activities of Daily Living (IADL)

• 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:
  • Valsartan 160 mg Bid: mean 0.71 versus 0.60 on control; difference 0.11 (95% CI -0.47 to 0.69), standardized effect d = 0.04; z-test p = 0.708.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Valsartan 160 mg Bid and the control. The interval gives the range of effects that remain compatible with the data.

The posted secondary results have a broadly similar pattern to the primary comparison: the available unadjusted estimates do not indicate a clear separation between LCZ696 and valsartan. The registry lists changes in cortical composite SUVr, individual cognitive domains, and the IADL summary score as secondary outcomes, but results are not posted for every listed outcome. The IADL comparison has an interval crossing zero and an unadjusted p-value of 0.708.

These secondary measures address related but distinct questions. Cognitive-domain measures can examine aspects of performance that a global composite may blend together, while cortical SUVr and IADL scores capture different types of outcomes. Agreement in the general lack of a clear group separation can provide context for the primary result, but it does not independently prove that there is no effect. Nor does a result on one secondary measure establish an effect on another domain.

Unadjusted secondary findings carry less weight because several outcomes and analyses create multiple opportunities to observe apparently noteworthy results by chance. The registry record does not report a multiplicity adjustment or a prespecified hierarchy for interpreting these outcomes. Without such information, secondary p-values should be treated as exploratory rather than as confirmatory evidence. The number of outcome measures and analyses in the record does not alone establish how the analysis plan handled multiplicity.

Secondary results also have outcome-specific analysis counts, and the record does not provide a complete account of missingness for each measure. Their interpretation therefore depends on both the statistical uncertainty and the completeness of reporting. They add breadth to the study’s outcome assessment, but they do not change the central limitation that the primary non-inferiority margin is not reported in the registry record.

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

• Change From Baseline in Cortical Composite Standardized Uptake Value Ratio (SUVr)
• Change From Baseline in Individual Cognitive Domains
• Change From Baseline in the Summary Score of the Instrumental Activities of Daily Living (IADL)

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is continuous: it is a score that can take a range of values, rather than a count of events. The arm summaries are least-squares means of change, accompanied by standard errors. A least-squares mean is a model-based estimate that accounts for specified covariates; it need not equal a simple arithmetic average. A standard error describes the estimated sampling uncertainty of an estimate, not the spread of individual participant scores.

What does the primary model do?

The registry describes a repeated-measures ANCOVA for CogState GCCS change. ANCOVA is a regression approach that compares groups while adjusting for measured baseline values and other recorded factors. Repeated-measures models use measurements from multiple visits and represent how observations from the same participant are related over time. The stated fixed-effect factors were treatment, age stratification factor, Mini mental state examination stratification factor, education level, Apolipoprotein E ε4 allele status, cerebrovascular disease burden at screening, visit, and treatment-by-visit interaction. The stated covariates were baseline GCCS and visit-by-baseline GCCS. A common unstructured covariance matrix among visits between treatment groups was used.

In general, a covariance structure describes how repeated measurements vary together. An unstructured covariance matrix allows the relationships among visits to be estimated without imposing a simpler pattern. This flexibility can represent varied correlations over time, but it also relies on the data supporting estimation of the model. The registry does not report alternative covariance structures or sensitivity analyses.

What does the missing-at-random assumption mean?

The primary analysis used a direct likelihood method with a missing-at-random assumption. In general, missing at random means that, conditional on information included in the model, the probability that an outcome is missing does not depend on the unobserved value itself. This is an assumption, not something proven by the analysis. The record does not report the amount or reasons for missing outcome data, nor does it describe sensitivity analyses under alternative missing-data assumptions.

How should the unadjusted recalculation be read?

The recalculated comparison is a simple, unadjusted comparison based on posted group summary statistics. It is useful for a straightforward numerical contrast, but it is not a reproduction of the sponsor’s repeated-measures ANCOVA. It does not adjust for the listed baseline factors or use the full repeated-visit model. Its confidence interval and p-value should therefore be interpreted as a separate summary calculation rather than as the primary model result. The registry’s own analysis is the more relevant result for understanding the analysis the study reports.

In general, a p-value describes how unusual data at least as incompatible with a specified null hypothesis would be under the statistical model. It is not the probability that the null hypothesis is true, and it does not measure the size or clinical importance of an effect. A p-value of 0.736 in the unadjusted recalculation is not evidence that the groups are identical; it indicates that the observed contrast is not unusual under a no-difference reference in that calculation.

What does the registry’s non-inferiority label establish?

Non-inferiority analyses compare a treatment effect with a prespecified margin that defines how much worse the experimental treatment could be while still meeting the study’s criterion. That margin and the direction of the formal testing rule are essential for interpreting a non-inferiority result. Although the reported primary analysis is labeled non-inferiority, the margin and its justification are not reported in the registry record. The label should not be taken by itself as proof that a non-inferiority criterion was met.

The registry reports an ANCOVA estimate and a 95% confidence interval for its treatment contrast. A confidence interval conveys a range of effect estimates compatible with the data and model at the stated confidence level. Whether that interval meets a non-inferiority criterion depends on the missing margin and the precise contrast convention. The interval crosses zero, which means it includes no difference, but that fact alone neither proves equality nor resolves a non-inferiority claim.

Finally, the analysis population is defined by receipt of at least one dose and non-missing primary values at both baseline and month 36. This is narrower than all randomized participants and can affect how broadly the result applies. The record does not provide a full accounting of participants with unavailable endpoint data or additional analysis populations, so the extent and implications of exclusions cannot be assessed from the registry record alone.

11. Confidence Intervals and Precision

The reported primary contrast has uncertainty that spans zero. The registry’s own 95% confidence interval crosses zero, and the unadjusted recalculated interval also crosses zero. In practical terms, neither interval identifies a direction of difference with high precision: values on either side of no difference remain compatible with the respective analysis. The posted estimates are close to zero relative to the uncertainty around them.

A confidence interval is a range generated by a statistical procedure that, under its assumptions, would cover the underlying effect at the stated rate over repeated samples. It is not a range containing a fixed proportion of individual participant outcomes, nor is it a probability statement about the particular effect after observing the data. The width reflects both the amount of information and variability, as well as modeling choices.

For the primary endpoint, the record reports a 95% interval for the adjusted ANCOVA contrast. The unadjusted recalculation uses an approximate pooled standard deviation derived from the posted standard errors, so its interval is a simple summary rather than the registry model’s interval. The registry record does not report a clinically important difference threshold or the non-inferiority margin. Consequently, the precision can be described statistically, but the interval cannot be judged against a clinical or non-inferiority threshold using this record alone.

Standard errors for the arm-level least-squares means reflect uncertainty in those estimates; they do not describe the range of cognitive scores among participants. The number of participants contributing to an endpoint is also relevant to precision, and the primary analysis counts are reported for participants with values at both baseline and month 36. Missing endpoint data and the missing-at-random assumption may affect precision and interpretation, but their impact cannot be quantified from the registry information.

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

The columns for 80% and 90% power describe the number of participants per arm required, under those assumptions, to have the stated chance of detecting the selected effect if that effect were the true one. Higher planned power requires more participants for a given effect. The rows that use 100%, 75%, and 50% of the observed effect illustrate how the target effect changes the planning requirement: smaller effects need more participants. For example, the primary illustration at 100% of the observed effect gives 13,042 participants per arm for 80% power and 17,459 participants per arm for 90% power. At 50% of the observed effect, the corresponding figures are 52,165 and 69,834 participants per arm.

These values are highly dependent on the observed effect, which is small in the posted primary comparison. They should not be read as a design requirement for this completed study or as evidence about the study’s original planning. This is not observed power: it is a forward-looking planning calculation using the observed effect as an input, not a measure of the probability that the completed study detected a true effect.

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 = 0.018; pooled standard deviation σ = 0.519.
2. n = 2 × (1.96 + 0.842)² × 0.519² / 0.018² = 13,041.1, rounded up to 13,042 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 <- 0.018                      # observed difference in means
sigma <- 0.518813359473409                     # 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(2884206)
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 = 13,042 (80% power), 17,459 (90% power)
 75% of observed effect: n per arm = 23,185 (80% power), 31,038 (90% power)
 50% of observed effect: n per arm = 52,165 (80% power), 69,834 (90% power)
power.t.test: 13042.1 
Simulated power at n = 13,042 per arm: 0.795 (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 registry’s own primary analysis is an ANCOVA labeled as non-inferiority, whereas the recalculated comparison is an unadjusted contrast derived from the posted arm summaries. The registry reports a negative adjusted treatment contrast and a 95% confidence interval that crosses zero. The simple recalculation, expressed as valsartan minus LCZ696, gives a positive mean difference of 0.018 and an unadjusted p-value of 0.736. The sponsor’s reported p-value is 0.736.

The different contrast direction does not by itself indicate a conflict in the findings: the contrast conventions differ, and the registry analysis adjusts for baseline and other stated factors while also modeling repeated visits. The estimates are close in magnitude and both indicate substantial uncertainty around a difference. The recalculation is useful as an independent descriptive check, but it does not replace the model specified in the registry record. Because the discrepancy indicator is false, no discrepancy is flagged between the posted analysis and the recalculation; interpretation should still center on the registry’s own analysis for the study’s inferential result.

The non-inferiority designation requires particular care. The record does not report the margin or its justification, so the sponsor’s analysis cannot be fully assessed against the formal criterion from the registry record alone. The model label and p-value do not supply that missing information.

15. Multiplicity, Missing Data and Analysis Populations

The registry record does not report whether the study controlled the overall type I error rate across the primary and secondary outcomes, how multiple analyses were prioritized, or whether any hierarchical testing procedure was prespecified. This information matters because testing multiple hypotheses can increase the chance of at least one apparently positive result even when there is no true effect. A correction or testing hierarchy can address that issue, but neither should be assumed when it is not described.

The record lists four outcome measures and six analyses, and three secondary outcomes are posted. The listed secondary outcomes span cognitive domains, cortical SUVr, and IADL. Their tests should not be treated as though each were the sole confirmatory endpoint unless the protocol’s testing strategy supports that interpretation. The original statistical analysis plan and any multiplicity procedures are not reported in the registry record.

Missing data are relevant to both the primary and secondary interpretations. The primary analysis population consists of Safety Set participants with non-missing GCCS values at baseline and month 36. The model used a direct likelihood method and assumed data were missing at random. In general, this assumption allows observed information in the model to contribute to estimation when outcomes are missing, but it does not guarantee that missingness has no effect. The record does not report the amounts or reasons for missing primary measurements, or sensitivity analyses under other assumptions.

The Safety Set included randomized patients who received at least one dose. That definition does not mean that every member contributed to the primary endpoint analysis, because the latter additionally required the two specified GCCS assessments. The registry record does not describe other primary analysis populations or provide a complete participant flow that would explain all differences in endpoint counts. As a result, the potential influence of missing observations and analysis-population choices cannot be fully evaluated.

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 2% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: cardiac failure (5.1%), atrial fibrillation (4.7%), acute myocardial infarction (4.0%), pneumonia (3.4%), chronic obstructive pulmonary disease (3.1%).

The registry lists serious adverse events by group, with at-risk counts and counts affected. During the randomized treatment groups, serious adverse events were reported for 145 of 295 at risk in the LCZ696 group and 161 of 297 at risk in the valsartan group. Deaths were recorded for 28 and 39 participants, respectively. These are descriptive counts from the registry record; they do not by themselves establish a causal difference between treatments.

The record also lists run-in groups, which are distinct from the randomized treatment groups. Those entries should not be combined with the randomized-group figures when interpreting treatment-period outcomes. The listed serious adverse-event terms include cardiac failure at 5.1%, atrial fibrillation at 4.7%, acute myocardial infarction at 4.0%, pneumonia at 3.4%, chronic obstructive pulmonary disease at 3.1%, and cardiac failure congestive at 2.7%.

These counts do not provide timing, exposure-adjusted rates, or a complete account of how events were adjudicated. The registry record also does not report a statistical comparison for these adverse-event figures. Group sizes, follow-up, and event ascertainment can affect interpretation, so the counts should be read descriptively rather than as a standalone comparative safety analysis.

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

The primary results describe the estimated change in a cognitive composite among participants with the required measurements and under the registry’s stated repeated-measures model. Both treatment groups had negative arm-level change estimates, and the adjusted and unadjusted comparisons have confidence intervals that cross zero. This indicates that the posted data do not precisely distinguish the groups on the primary measure.

That finding is not proof that the treatments have identical effects. A lack of a clear difference can arise when the true effects are similar, when the data are imprecise, or when the analysis cannot resolve smaller differences. The registry’s primary analysis is labeled non-inferiority, but the non-inferiority margin is not reported in the registry record. Without the margin and its justification, the formal non-inferiority claim cannot be fully evaluated from this record.

The endpoint is a standardized cognitive test composite, not a complete measure of cognitive health or daily life. The registry also lists IADL, cognitive-domain, and cortical SUVr outcomes, but not all secondary results are posted. The findings should therefore be understood as specific to the outcomes, follow-up, and analysis population recorded. The primary analysis required non-missing GCCS assessments at baseline and month 36, and the model assumed missing at random.

The results do not establish that either treatment improves or worsens cognition, and they do not establish a treatment effect outside the population and study conditions described. They also do not answer questions about outcomes that were not measured or not reported in the registry record. The adverse-event entries provide descriptive counts but do not constitute a complete causal or exposure-adjusted safety comparison.

Overall, the most defensible reading is that the recorded primary analysis did not clearly separate the treatments on the measured cognitive composite, while important details needed to assess the non-inferiority criterion and missing-data sensitivity are absent from the registry record.

18. Why This Trial Matters Statistically

This trial is statistically informative because it uses a continuous cognitive composite as its primary outcome and evaluates change over repeated visits. That setting illustrates why the reported model matters: a repeated-measures ANCOVA can use information across visits and adjust for prespecified baseline factors, while a simple contrast of arm summaries answers a narrower, unadjusted question. The two should not be treated as interchangeable estimates.

The result also demonstrates the distinction between absence of a clearly detected difference and evidence that two treatments are equivalent or non-inferior. A confidence interval crossing zero is not sufficient to establish either conclusion. A non-inferiority interpretation depends on a margin, its direction, and a testing plan. Because the margin and its justification are not reported in the registry record, the analysis label cannot on its own settle the substantive claim.

Missing data and analysis populations are central to interpreting longitudinal cognitive outcomes. The primary population excludes participants without both baseline and month 36 GCCS values, and the model assumes missing at random. The record does not report enough information to assess the strength of that assumption or the robustness of the result to alternatives.

Finally, the mix of primary and secondary outcomes highlights the role of multiplicity and selective interpretation. Secondary measures can broaden understanding, but unadjusted analyses across multiple outcomes have less confirmatory weight when the testing strategy is not reported. Together, these features make the trial a useful case for reading estimates alongside their model, interval, assumptions, and reporting gaps rather than relying on a single p-value or label.

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.

Several trial-specific details limit interpretation. The registry labels the primary analysis as non-inferiority, but the margin, its justification, and the formal decision rule are not reported in the registry record. Without these, it is not possible to assess whether the confidence interval meets the study’s intended non-inferiority criterion.

The primary endpoint analysis includes Safety Set participants with non-missing GCCS at baseline and month 36, rather than all randomized participants. The record does not give a detailed accounting of missing outcome measurements or their reasons. Although the primary model assumes missing at random, sensitivity analyses assessing other missingness assumptions are not reported in the registry record.

The registry lists secondary outcomes beyond the primary composite, but results are not posted for every listed outcome. A multiplicity strategy and outcome hierarchy are also not reported. This limits the ability to judge how much evidential weight to place on secondary comparisons or whether their interpretation was prespecified.

The run-in period selected for tolerability before randomization, according to the arm descriptions. The record does not provide detailed eligibility criteria or enough information to assess how that selection affects generalizability. Safety results include run-in and randomized treatment groups, and the registry does not provide exposure-adjusted comparisons or a statistical analysis of the adverse-event figures.

Finally, the registry record does not report the original sample-size justification or interim-analysis plan. Planning illustrations based on the observed effect cannot fill those gaps and should not be used to infer the original design rationale.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Efficacy and Safety of LCZ696 Compared to Valsartan on Cognitive Function in Patients With Chronic Heart Failure and Preserved Ejection Fraction — NCT02884206.
• Registry record last updated 2024-08-06; status last verified 2024-07.
• Results first posted 2024-08-06.
• 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

NCT02884206 is a completed phase 3, randomized, double-blind, parallel-group, active-controlled study of LCZ696 compared with valsartan in patients with chronic heart failure and preserved ejection fraction. Enrollment was 592, and the primary endpoint was change from baseline to month 36 in CogState GCCS.

The registry’s primary ANCOVA analysis is labeled non-inferiority. The reported interval crosses zero, and the unadjusted recalculation also does not show a clear difference. The non-inferiority margin is not reported in the registry record, limiting formal interpretation of that label. The primary model assumed missing at random, and its analysis population required GCCS values at both baseline and month 36.

Secondary outcomes and adverse-event counts provide additional descriptive context, but multiplicity procedures, detailed missing-data information, and some outcome results are not reported in the registry record.