Study of Verinurad in Heart Failure With Preserved Ejection Fraction: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Study of Verinurad in Heart Failure With Preserved Ejection Fraction (NCT04327024), a randomized phase 2 study. The primary outcome is Change From Baseline at Week 32 in Peak V02 Consumption in Verinurad + Allopurinol Compared to Placebo (ANCOVA Model).

2. Clinical Question

The clinical question is about a measured exercise-capacity outcome, not a direct measure of every aspect of heart failure or of clinical events. The primary endpoint is a change score at a specified follow-up, and the reported contrast estimates the difference between the two randomized groups after accounting for baseline peak VO2 in the stated model. The registry record also lists secondary questions about peak VO2 compared with allopurinol and about the Kansas City Cardiomyopathy Questionnaire total symptom score (KCCQ-TSS). The secondary outcome results are not reported in the registry record.

3. Trial Design

This was a randomized, parallel-group Phase 2 study with 159 participants enrolled. The registry lists the study as completed and gives a start date of 2020-05-19 and a primary completion date of 2022-04-29. Its condition is heart failure with preserved ejection fraction. AstraZeneca is listed as the sponsor.

The groups were verinurad plus allopurinol, allopurinol alone, and placebo. Each group had 53 participants in the primary comparison record. The combination group received titration steps leading to a target dose of 12 mg verinurad and 300 mg allopurinol. The allopurinol-alone group received allopurinol titration to 300 mg, while placebo was described in three steps. The record lists interventions for both active drugs and their corresponding placebos.

The registry identifies the allocation as randomized and the model as parallel. It lists masking as triple, while the official title describes the study as double-blind. The record does not explain this difference, so the masking arrangements should not be inferred beyond those descriptions. It describes the study as multicentre and as including placebo and active control.

The primary endpoint was change from baseline at Week 32 in peak VO2 for verinurad plus allopurinol compared with placebo. The primary analysis used the Full Analysis Set and ANCOVA. The registry describes a hierarchical testing sequence covering the primary and secondary objectives to address multiple testing and control the overall two-sided Type I error rate at 0.05. Since the primary endpoint was not rejected at a two-sided 0.05 level, the sequence did not continue. Further details of the protocol, including the original sample-size justification and any interim analyses, 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: Subjects who met all the of inclusion criteria and none of the exclusion criteria were randomized to a study treatment. Study treatment was titrated over 8 weeks. Colchicine prophylaxis was given during the titration period and during the first 4 weeks of treatment at target dose (12 weeks total) and was distributed as available, currently 500 μg within the European Union (EU) and 600 μg within the United States (US).

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: The Full Analysis Set includes all subjects who have been randomized to study treatment.

6. Primary Endpoint

The registry lists the primary outcome as Change From Baseline at Week 32 in Peak V02 Consumption in Verinurad + Allopurinol Compared to Placebo (ANCOVA Model), measured over From baseline to Week 32.

Peak VO2 is the peak rate of oxygen consumption measured during exercise, expressed in this study in mL/kg/min. It is used as a continuous measure of exercise capacity. The primary outcome was not the Week 32 value alone: it was the change from baseline to Week 32, defined in the registry as the Week 32 value minus the baseline value. The comparison asks whether the average change differed between the verinurad-plus-allopurinol group and placebo.

In general, a continuous endpoint retains information about the measured value across participants rather than reducing it to a binary outcome such as improved or not improved. An analysis that includes baseline peak VO2 as a covariate can account for baseline differences in the model while estimating the treatment-group contrast. That adjustment does not change the endpoint into a different measure; the outcome remains change in peak VO2.

Peak VO2 offers a defined physiological measure, but it does not by itself capture all symptoms, quality of life, hospital events or other outcomes relevant to HFpEF. The registry lists KCCQ-TSS and peak VO2 comparisons as secondary outcomes. Results for those secondary outcomes are not reported in the registry record, so they cannot expand the interpretation of the primary result.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does the ANCOVA estimate?

The registry describes an ANCOVA using change from baseline in peak VO2 as the dependent variable, treatment as the independent variable and baseline peak VO2 as a covariate. The primary comparison is verinurad plus allopurinol against placebo. The reported estimate is a mean difference of -0.10 mL/kg/min; the 95% confidence interval is -1.28 to 1.08 and the p-value is 0.862.

In general, ANCOVA estimates a group contrast while accounting for a covariate included in the model. In this analysis the covariate is baseline peak VO2. This can improve the comparison's precision when baseline values are related to follow-up outcomes, but the registry does not report the model diagnostics or the extent of that relationship. The model is described in terms of the change score as outcome and baseline peak VO2 as covariate; further specification, such as additional covariates, is not reported in the registry record.

How should the confidence interval be read?

A confidence interval expresses uncertainty around an estimated contrast under the analysis model. In general, a 95% confidence interval is generated by a procedure that, under its assumptions and repeated use, captures the true parameter at the stated rate. It is not a statement that there is a 95% probability that this particular fixed parameter lies inside this particular interval. For this trial, the reported interval ranges from -1.28 to 1.08 mL/kg/min and includes zero. Its width means the data do not pin down a single narrow estimate of the difference.

The p-value of 0.862 is a measure of how unusual the observed test statistic, or a more extreme one, would be under the null hypothesis of no mean difference and the model assumptions. It is not the probability that the null hypothesis is true, and it is not a measure of the size or clinical importance of a possible difference. The registry specifies a two-sided null hypothesis of zero difference and a two-sided alternative that the difference is not zero.

Why use a hierarchical testing sequence?

The registry says a hierarchical sequence was used for confirmatory analyses of the primary and secondary objectives to address multiple testing and control the overall two-sided Type I error rate at 0.05. In general, testing several hypotheses can increase the chance of at least one false-positive result if each is treated as an independent opportunity for confirmation without an adjustment strategy. A prespecified sequence can control that risk by requiring tests to pass in order before proceeding to later claims.

Here, peak VO2 compared with placebo was first. Since this endpoint was not rejected at a two-sided 0.05 level, the sequence did not continue. This explains why later outcomes in the hierarchy cannot be interpreted as having passed confirmatory testing on the basis of this sequence. The precise ordering beyond the primary test and any additional details of the procedure are not reported in the registry record.

What does the analysis population mean?

The primary analysis population is named the Full Analysis Set. The registry does not describe the exact membership rules for that set or how departures from randomized assignment were handled. The reported comparison includes 53 participants in each of the two primary groups. The record does not give endpoint-specific missingness, the number with observed Week 32 measurements, or the method used to handle missing outcome data. Those details matter because an analysis of observed outcomes can be affected if missingness is related to outcomes or treatment assignment.

In general, randomization supports a fair comparison by balancing known and unknown factors on average, while an analysis population and missing-data approach determine how the observed data contribute to the estimate. The available registry description identifies the population label and model but does not provide enough detail to assess all assumptions about missing data or adherence.

What does the planning illustration represent?

The sample-size assumptions describe a planning illustration based on the observed effect, a two-sided significance level of 0.05, equal allocation between two arms and a normal approximation. The effect to detect is described as a fraction of the observed effect. In general, such calculations illustrate the sample size that might be required under stated assumptions; they are not observed power and do not reveal the original design justification. The registry record does not report the trial's own sample-size justification. No planning estimate should be mistaken for an observed result or for evidence about the probability that this trial detected an effect.

10. Analyses Reported by the Sponsor

The registry's reported primary analysis is an ANCOVA comparison with a mean difference of -0.10 mL/kg/min, a 95% confidence interval from -1.28 to 1.08 and p = 0.862. This is the sponsor-reported inferential result for the primary endpoint and is the principal contrast to use when describing the trial's posted findings.

The group estimates are 0.27 mL/kg/min for verinurad plus allopurinol and 0.37 mL/kg/min for placebo. They are reported as least-squares means with their own confidence intervals. A comparison reconstructed from group summaries would be an unadjusted calculation and would not necessarily reproduce the ANCOVA contrast, which includes baseline peak VO2 as a covariate. The registry record does not report a separate recalculated contrast for this endpoint.

No discrepancy between the sponsor analysis and a recalculation is flagged in the registry record. There is therefore no stated conflict to resolve. The reported ANCOVA result should be distinguished from any simple, unadjusted calculation based on the group summaries, because they need not estimate the same quantity under the same model.

11. Multiplicity, Missing Data and Analysis Populations

The record explicitly describes a hierarchical confirmatory testing sequence for the primary and secondary objectives. Its stated purpose was to address multiple testing and control the overall two-sided Type I error rate at 0.05. The primary Week 32 peak VO2 comparison against placebo came first. Because it was not rejected at the two-sided 0.05 level, the registry says the sequence did not continue.

Three secondary outcomes are listed: peak VO2 compared with allopurinol, KCCQ-TSS compared with placebo, and KCCQ-TSS compared with allopurinol. Their results are not reported in the registry record. Their listing alone does not establish a statistically supported effect, and the stated stopping rule means they were not advanced through the confirmatory sequence after the primary result.

The analysis population for the primary endpoint is the Full Analysis Set. The registry does not provide the operational definition of that set, explain exclusions, or report endpoint-specific counts of missing baseline or Week 32 measurements. The method for handling missing values, the assumptions underlying it and sensitivity analyses are not reported in the registry record. These omissions limit assessment of how missing outcome data may have affected the estimate.

Further details of the hierarchical order and any interim analyses are not reported in the registry record. The sample-size rationale is also not reported. A planning illustration based on the observed effect and stated assumptions is not a substitute for the original design justification and is not observed power. These missing details limit reconstruction of the full planned analysis strategy, but they do not change the primary estimate and test as posted.

12. 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 3% 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 (9.4%), acute myocardial infarction (3.8%), atrial fibrillation (3.8%), pneumonia (3.8%), cardiac failure acute (1.9%).

The registry reports serious adverse events for 10 of 53 participants in the verinurad-plus-allopurinol group, 10 of 53 in the allopurinol group and 9 of 53 in the placebo group. It separately reports other adverse events for 24 of 53, 26 of 53 and 20 of 53 participants, respectively. Deaths are listed as 1, 2 and 1 across those same groups. These are descriptive counts in the registry; they are not the primary efficacy endpoint and should not be treated as a randomized statistical comparison unless an analysis is reported.

The record lists serious-event terms with percentages, including cardiac failure at 9.4%, acute myocardial infarction at 3.8%, atrial fibrillation at 3.8%, pneumonia at 3.8%, cardiac failure acute at 1.9% and breast cancer at 1.9%. The record does not provide enough detail in the listed information to interpret these percentages as comparative risks by treatment group. Event counts alone do not account for differences in exposure time, event recurrence or uncertainty around the estimates.

These safety figures are limited to the categories and counts posted in the registry. The record does not provide a detailed causal assessment in this information, so the counts should not be read as proof that a study treatment caused or prevented an event. The listed deaths are included as counts, not as evidence of a treatment effect.

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

The primary result is an estimated mean difference in Week 32 change in peak VO2 between verinurad plus allopurinol and placebo. The reported contrast is -0.10 mL/kg/min, with a 95% confidence interval from -1.28 to 1.08 and p = 0.862. Under the specified ANCOVA, the result does not reject the null hypothesis of no mean difference at the stated two-sided 0.05 threshold. The interval includes zero and is compatible with effects in either direction.

This is not proof that the two groups are identical. A failure to reject a null hypothesis can occur when the data do not distinguish the estimate from zero with sufficient precision, and the confidence interval still includes a range of possible differences. The registry does not provide a threshold for a clinically important difference in peak VO2, so the result cannot be classified against such a threshold using the recorded information.

The inference applies to the primary endpoint, the comparison against placebo, the Week 32 time point and the Full Analysis Set as named in the record. It does not directly answer whether the treatment groups differed on all symptoms, quality of life, hospitalization, survival or other clinical outcomes. The listed secondary outcomes concern peak VO2 and KCCQ-TSS, but their results are not reported in the registry record and the confirmatory sequence stopped after the primary test.

The adverse-event counts describe what the registry records for the study groups, but they are not evidence of causation or a substitute for a full comparative safety analysis. The study's Phase 2 designation and enrollment of 159 participants provide context, not an independent basis for concluding that an effect was present or absent. The result should be understood as an uncertain estimate for one defined measure and comparison, rather than a universal statement about treatment or HFpEF.

14. Why This Trial Matters Statistically

This trial illustrates why interpretation should focus on the contrast and its uncertainty rather than on whether group means appear different. The primary analysis reports a mean difference of -0.10 mL/kg/min and a 95% confidence interval from -1.28 to 1.08. The interval crosses zero, and the p-value is 0.862. Taken together, these statistics show that the posted analysis does not establish a difference on the primary endpoint, while also leaving uncertainty about the size and direction of the underlying contrast.

It also shows the distinction between group-level estimates and an adjusted comparison. The primary model used change from baseline as the outcome and baseline peak VO2 as a covariate. The reported ANCOVA contrast is therefore the relevant modeled comparison, rather than an informal comparison of the two group estimates alone. Understanding the estimand and model is essential when interpreting clinical trial results.

The hierarchical plan adds another statistical lesson. A trial can list several outcomes, but the order and rules for testing determine which results can support confirmatory claims. Here, the sequence ended when the primary endpoint was not rejected at the specified threshold. The secondary outcomes cannot be presented as if they independently passed the planned confirmatory procedure.

Finally, the record distinguishes what is described from what remains unavailable. It names the Full Analysis Set and the model, but does not give details of missing-data handling or the trial's sample-size justification. Transparent reporting of those elements helps readers judge how robust an estimate is and how closely the reported analysis corresponds to the planned design.

15. 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 supplies a concise description of the primary model, but not a full statistical analysis plan. It does not report model diagnostics, the exact handling of missing Week 32 values, sensitivity analyses, or the operational definition of the Full Analysis Set. Those omissions constrain assessment of the assumptions behind the estimate and whether alternative reasonable analyses would yield similar findings.

The record lists 159 enrolled participants and 53 participants in each of the three groups for the reported group summaries. It does not provide the number of participants with observed primary endpoint data at Week 32 separately from the reported analysis counts. Nor does it report reasons for missing data or withdrawals in the outcome information. The degree to which missingness could affect the analysis therefore cannot be assessed from the recorded details.

The study is Phase 2 and the primary outcome is a continuous exercise measure at Week 32. The result does not by itself describe longer-term outcomes or other dimensions of HFpEF. The registry lists secondary peak VO2 and KCCQ-TSS outcomes, but their results are not reported in the registry record. The confirmatory sequence also stopped after the primary test was not rejected.

The record lists masking as triple, while the official title describes the study as double-blind; it does not explain the difference. Details of the sample-size rationale and any interim analyses are not reported in the registry record. Finally, the serious adverse-event and death counts are descriptive in the posted information, without a reported comparative inferential analysis or additional context sufficient to establish causation.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Study of Verinurad in Heart Failure With Preserved Ejection Fraction — NCT04327024.
• Registry record last updated 2023-06-29; status last verified 2023-06.
• Results first posted 2023-06-29.
• 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.

19. Record Summary

NCT04327024 is a completed randomized, parallel-group Phase 2 study of verinurad with allopurinol in HFpEF. It enrolled 159 participants and compared the combination with allopurinol alone and placebo. The primary outcome was change from baseline to Week 32 in peak VO2, analysed in the Full Analysis Set using ANCOVA with baseline peak VO2 as a covariate.

For the primary comparison with placebo, the reported mean difference was -0.10 mL/kg/min, with a 95% confidence interval from -1.28 to 1.08 and p = 0.862. The registry says the primary endpoint was not rejected at the two-sided 0.05 level, so the hierarchical confirmatory testing sequence did not continue. Results for the listed secondary outcomes are not reported in the registry record.

The registry also reports group-level adverse-event and death counts. These are descriptive and do not, by themselves, establish comparative treatment effects or causation. Further details of missing-data handling, the sample-size justification and any interim analyses are not reported in the registry record.