Phase 2 Healthy Volunteer Study to Evaluate the Ability of PRT064445 to Reverse the Effects of Several Blood Thinner Drugs on Laboratory Tests (Module 1 of 4): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Phase 2 Healthy Volunteer Study to Evaluate the Ability of PRT064445 to Reverse the Effects of Several Blood Thinner Drugs on Laboratory Tests (Module 1 of 4) (NCT01758432), a randomized phase 2 study. The primary outcome is Efficacy: Percent Change From Baseline in Anti-fXa Activity at 2 Mins Following Andexanet/Placebo Administration.

2. Clinical Question

The posted endpoint concerns a brief measurement window: baseline to two minutes after administration ended. The study also listed thrombin generation and unbound apixaban concentration as secondary efficacy outcomes, alongside pharmacokinetic measures of PRT064445. Those additional measures provide context for the laboratory and drug-exposure questions, but the reported primary endpoint is specifically anti-fXa activity.

The central comparison is therefore about the size and precision of short-term laboratory changes in healthy volunteers. It is not a direct comparison of clinical outcomes, and the registry record does not report evidence about patient-centered outcomes such as bleeding or survival.

3. Trial Design

This was a randomized, phase 2, parallel-group study with quadruple masking, described in the official title as double-blind and vehicle-controlled. It was conducted in healthy volunteers and was marked completed in the registry. The record lists enrollment as 54. The start and primary completion dates are recorded as 2012-12 and 2015-09.

The trial title identifies Module 1 as one part of a four-module program and describes assessment of intravenous PRT064445 after dosing to steady state with one of four direct or indirect fXa inhibitors. The Module 1 intervention labels refer to PRT064445 with apixaban, placebo with apixaban, and placebo. The exact apixaban schedule and the allocation of participants among all background-drug modules are not reported in the registry record.

For the primary endpoint, the registry reports a placebo group of 18 and PRT064445 regimen groups of 6 each. The listed active regimens were a 90 mg bolus, a 210 mg bolus, a 420 mg bolus, a 420 mg bolus followed by a 180 mg infusion, a 420 mg bolus followed by a 180 mg bolus, and a 420 mg bolus followed by a 480 mg infusion. The detailed arm descriptions specify intravenous administration; some descriptions also state infusion rates and durations.

The registry describes the primary endpoint population as subjects who received andexanet or placebo and were included in the pharmacodynamics analysis. It does not report a full account of exclusions, missing measurements, or the handling of incomplete endpoint data. A rationale for the enrollment target, a formal sample-size justification, and details of any interim analyses are not reported in the registry record.

Randomization and masking are features of the design, but their presence alone does not establish that every possible source of bias was controlled. Interpretation also depends on how the endpoint was measured, which participants contributed data, how comparisons were modeled, and how multiple planned outcomes were handled. Some of these details are available in the endpoint description and sponsor analyses; others 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: Apixaban was administered orally at 5 mg twice daily for 6 days to steady-state in an open label fashion. Subjects were then randomized to receive study treatment (andexanet:placebo, 6:3) intravenously at different doses/dose regimens on Day 6, all bolus doses administered such that they ended at 3 hours after the last dose of apixaban.

Only the first 6 of 7 groups are shown.

Overall Study

Analysis population for the primary outcome, as defined in the registry: 54 subjects who received andexanet or placebo were included in the pharmacodynamics (PD) analysis

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: 54 subjects were enrolled in this study.

Only the first 6 of 8 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as Efficacy: Percent Change From Baseline in Anti-fXa Activity at 2 Mins Following Andexanet/Placebo Administration, measured over Baseline to 2 minutes following the end of andexanet/placebo administration.

The primary endpoint is percent change from baseline in anti-fXa activity, measured two minutes after the end of andexanet or placebo administration. The registry says that anti-fXa activity was measured immediately before administration for baseline and again at the stated follow-up time, using a commercial kit identified in the record. The endpoint is reported as a mean with standard deviation.

Anti-fXa activity is a laboratory measure related to the activity of fXa inhibitors. In this study, measuring the change soon after administration allowed investigators to assess a short-term pharmacodynamic response under the study conditions. A percent-change outcome expresses the post-administration measurement relative to baseline, rather than reporting only the post-administration value. The registry does not give the calculation formula or further details about assay validation, measurement variability, or how values below a measurement limit would be handled.

The arm-level means are negative: for example, the placebo mean is -7.06, while the active-regimen means range from -67.79 to -95.04. These are group averages, not results for every individual. The associated standard deviations describe variation among observations within each group; they are not confidence intervals for the group means.

This endpoint is useful for assessing a laboratory response, but it is not itself a clinical outcome. The measurement at two minutes does not establish how long a response lasts, whether it changes bleeding, or whether it predicts outcomes in patients. Those questions are not answered by the primary endpoint as reported.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is continuous: each participant’s measurement is expressed as a percent change, and the registry reports a group mean and standard deviation. A mean describes the central value of a set of measurements. A standard deviation describes how spread out individual measurements are around that mean. It is not the same as a confidence interval, and it does not directly express uncertainty in the estimated mean or in a comparison between groups.

What does the ANCOVA comparison represent?

The sponsor used analysis of covariance, or ANCOVA, for the reported primary endpoint comparisons. In general, ANCOVA compares group outcomes while accounting for one or more covariates specified in the model. This can produce adjusted group estimates, often called least-squares means, and an adjusted difference between groups. Adjustment can be useful when the analysis plan identifies relevant baseline variables or design factors. The registry record does not report the full ANCOVA specification, including the covariates, model diagnostics, or any transformations. The adjusted differences should therefore be interpreted as the sponsor’s reported model-based comparisons, not as simple subtractions of the displayed arm means.

The reported values are labeled “Least Squares Means (Difference),” and the comparisons are labeled superiority. The registry lists p < 0.001 for each reported comparison. It does not report confidence-interval bounds, although the analysis entries identify a 95% confidence level. A confidence interval is a range produced by a statistical procedure to express uncertainty around an estimated quantity. Without its bounds, readers cannot assess the reported range of plausible adjusted differences directly. A p-value alone does not provide that range or indicate whether the difference is important in clinical practice.

Why do the adjusted values differ from the arm means?

The arm means are descriptive summaries. An adjusted ANCOVA contrast is a model-based estimate that can reflect covariate adjustment and the model’s definitions of group estimates and contrasts. The contrast’s sign depends on the order used for subtraction, and the registry record does not make that order clear enough to assign a directional interpretation to the reported negative values. It is therefore important not to treat the adjusted difference as a raw difference between the listed means or infer the subtraction order from its sign alone.

What does variation within groups tell us?

The standard deviations are notably different across the posted groups. In general, unequal variability can matter because methods that assume a common variance may not describe all groups equally well. When a continuous outcome has more variability in some groups than others, standard errors and tests can be affected if the model’s variance assumptions are unsuitable. That concern is especially relevant when group sizes are small. The registry record does not report diagnostic checks or a sensitivity analysis addressing unequal variances, so it is not possible to determine from the record whether this issue affected the conclusions.

General statistical interpretation also distinguishes a group average from an individual response. A mean can summarize a consistent shift, but it can also conceal variation among participants. The standard deviation offers information about that variation, though the registry does not provide participant-level data or the distribution of individual changes. Consequently, the means and standard deviations do not show how many individuals had a response of a particular size.

How should the p-values be read?

A p-value evaluates compatibility between observed data and a specified null model, subject to the model’s assumptions. The reported p < 0.001 values indicate strong statistical evidence against the null hypothesis used for those comparisons. They do not give the probability that the null is true, the probability that a result will replicate, or the magnitude of a clinically meaningful effect. Those interpretations require other information, including estimates, uncertainty intervals, study context, and appropriate handling of the planned comparisons.

Multiplicity is relevant because the registry lists multiple dose-versus-placebo comparisons and several secondary outcomes. When many hypotheses are tested, the chance of at least one apparently unusual result can increase unless the analysis strategy accounts for the testing structure. The registry record does not report whether or how multiplicity was controlled. The listed p-values should be read as reported, without assuming a particular adjustment procedure.

What can planning calculations say?

General sample-size calculations estimate how many participants might be needed for a future study under stated assumptions about the target effect, variability, significance level, allocation, and statistical method. The planning assumptions described for the posted illustrations use a two-sided significance level of 0.05, equal allocation, a normal approximation, and fractions of the observed effect. Such illustrations are planning scenarios, not observed power and not evidence that the original study was adequately powered for every endpoint. The trial’s own sample-size justification is not reported in the registry record.

For a continuous endpoint, a planning result expressed as participants per arm is not the same quantity as the total number of events used in some time-to-event designs. The distinction matters when reading sample-size estimates. No planning estimate should be treated as a measured trial result or as confirmation of the study’s design rationale.

Finally, statistical significance is only one component of interpretation. The endpoint is an early laboratory measurement, the population is healthy volunteers, the active groups are small, and confidence-interval bounds are absent from the reported comparisons. These features constrain conclusions about precision, generalizability, and clinical relevance even when the reported p-values are small.

10. Confidence Intervals for Each Arm

No two-arm comparison is shown for this trial, but each arm's mean can be given with a t-based 95% confidence interval computed from the posted mean, spread and number of participants.

R code and output

arm <- c("Module 1 Placebo", "Module 1 ( 90 mg)", "Module 1 (210 mg)", "Module 1 (420 mg)", "Module 1 (420 mg Bolus + 180 mg Infusion)", "Module 1 (420 mg Bolus + 180 mg Bolus)")
n   <- c(18, 6, 6, 6, 6, 6)
m   <- c(-7.06, -67.79, -78.51, -95.04, -94.03, -93.07)
sp  <- c(10.801, 7.731, 3.815, 1.375, 1.87, 1.765)
sd  <- sp      # posted standard deviations
for (i in seq_along(n)) {
  se <- sd[i] / sqrt(n[i]); h <- qt(0.975, n[i] - 1) * se
  cat(sprintf("CARM|%s|%d|%.3f|%.3f|%.3f|%.3f\n", arm[i], n[i], m[i], sd[i], m[i] - h, m[i] + h))
}

CARM  Module 1 Placebo  18  -7.060  10.801  -12.431  -1.689
CARM  Module 1 ( 90 mg)  6  -67.790  7.731  -75.903  -59.677
CARM  Module 1 (210 mg)  6  -78.510  3.815  -82.514  -74.506
CARM  Module 1 (420 mg)  6  -95.040  1.375  -96.483  -93.597
CARM  Module 1 (420 mg Bolus + 180 mg Infusion)  6  -94.030  1.870  -95.992  -92.068
CARM  Module 1 (420 mg Bolus + 180 mg Bolus)  6  -93.070  1.765  -94.922  -91.218

11. Analyses Reported by the Sponsor

The registry’s sponsor-reported primary comparisons use ANCOVA and report least-squares-mean differences, with p < 0.001 for each listed comparison. The arm summaries separately report unadjusted means and standard deviations. These are different statistical quantities: the arm summaries describe the observed group distributions, while the ANCOVA estimates are adjusted comparisons under the sponsor’s model.

The registry record does not include a separate unadjusted recalculated contrast alongside the sponsor’s ANCOVA results. Nor does it provide the full model specification or confidence-interval bounds needed to independently reconstruct and assess the adjusted comparisons. The reported estimates should therefore be described as the sponsor’s own analysis, rather than replaced by arithmetic based on the arm means.

The record identifies the ANCOVA results as superiority analyses. Because the contrast coding is not fully described, the negative signs should not be assigned a direction by assuming a particular subtraction order. The arm means do support a straightforward descriptive statement: the reported mean percent changes are more negative in each active-regimen group than in the placebo group. That observation is distinct from the adjusted model result.

The sponsor’s p-values provide evidence against the model’s null hypothesis for the reported comparisons, but they do not resolve questions about clinical meaning, precision, model assumptions, or multiplicity. Those questions require information that is not reported in the registry record.

12. Multiplicity, Missing Data and Analysis Populations

The registry lists one primary endpoint, eight posted secondary outcomes, and 18 analyses. The secondary outcomes include thrombin generation, unbound apixaban concentration, and pharmacokinetic measures of PRT064445. The primary endpoint itself has multiple dose-regimen comparisons with placebo. This creates a setting in which multiplicity is relevant: when several hypotheses are evaluated, the chance of observing at least one small p-value under a collection of null hypotheses can be greater than for a single test.

The registry record does not report a prespecified hierarchy among the comparisons, a family of hypotheses, or a multiplicity-adjustment method. It is therefore not possible to determine whether the reported p-values were adjusted, whether they are intended as confirmatory or exploratory, or how the secondary outcomes were incorporated into an overall testing strategy. The p-values should be read as the values reported for the individual sponsor analyses, not as evidence that any unreported multiplicity procedure was used.

The endpoint description identifies 54 subjects in the pharmacodynamics analysis. The record does not provide participant flow by analysis, the number with missing endpoint measurements, reasons for missingness, or methods used to handle incomplete data. It also does not report whether the analysis followed an intention-to-treat principle or another prespecified population definition. The phrase “included in the pharmacodynamics analysis” identifies the reported analysis population, but it does not by itself explain how that population was formed.

Because the endpoint is measured at baseline and shortly after administration, missingness at either time could affect calculation of a participant’s change. General methods for handling missing data make different assumptions about why values are unavailable; the appropriate interpretation depends on the missingness pattern and the analysis plan. Neither is reported in the registry record. There is also no reported sensitivity analysis showing whether conclusions change under alternative missing-data assumptions.

These omissions do not negate the posted results, but they limit independent assessment of how robust the statistical conclusions are to multiplicity and incomplete data. The reported group summaries and sponsor comparisons remain interpretable as recorded, while the broader inferential framework is not fully documented.

13. 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 3% reporting threshold.

Most frequent non-serious adverse event terms

The registry reports 0 serious adverse events in the placebo group and 0 in each listed PRT064445 group. The number at risk is reported as 18 for placebo and 6 for each active regimen. These are the registry’s group-level serious-event figures; they do not establish that the study had enough information to rule out uncommon harms or to characterize safety in populations beyond the healthy volunteers studied.

The record also reports other, non-serious adverse events by group: 14 in placebo, 3 in the 90 mg group, 3 in the 210 mg group, 5 in the 420 mg group, 4 in the 420 mg bolus plus 180 mg infusion group, 2 in the 420 mg bolus plus 180 mg bolus group, and 3 in the 420 mg bolus plus 480 mg infusion group. These are counts of participants affected as recorded, not percentages or rates. No group totals or percentages are calculated.

Death counts are not reported in the registry record. The record does not provide the specific non-serious event terms in the information summarized for these groups, so the counts cannot identify the type, severity, timing, or duration of those events. The absence of reported serious events in these groups should be read narrowly as the observed registry result, not as proof that serious events are impossible or that the intervention is safe in other settings.

Safety interpretation also depends on follow-up duration, event definitions, collection procedures, and the population studied. Those details are not all reported in the registry record. These figures should be understood as descriptive reporting from this study rather than as a comprehensive assessment of risk.

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

The results show that mean anti-fXa activity changed from baseline shortly after administration in the studied groups. The recorded mean percent changes were more negative in the PRT064445 regimen groups than in the placebo group. The sponsor’s reported ANCOVA comparisons were labeled superiority analyses and had p < 0.001. Together, these findings describe a short-term laboratory response under the trial conditions.

They do not establish that PRT064445 prevents or stops bleeding, improves survival, or changes another clinical outcome. The trial enrolled healthy volunteers, and the primary endpoint was a laboratory measurement taken two minutes after the end of administration. A laboratory response can be relevant to pharmacology without proving that a patient outcome will improve. The registry record does not report patient clinical-outcome results for this endpoint.

The findings also do not establish the duration of the measured response. The primary endpoint timeframe is limited to baseline and the specified post-administration measurement. Although other secondary measures are listed, their numeric results are not included in the registry record described for the primary analysis. The record therefore does not support conclusions about persistence, later rebound, or how the primary change relates to other measurements.

Nor do the results identify a best regimen. The listed group means can be compared descriptively, but the registry does not report a direct statistical comparison among active regimens. The sponsor’s reported analyses compare each listed regimen with placebo, and the record does not provide confidence-interval bounds for those estimates. Differences among the active-regimen means should not be treated as proof of dose superiority or equivalence.

Finally, results from healthy volunteers do not automatically generalize to patients or to other clinical settings. The record describes Module 1 within a broader four-module study, but the information reported for this analysis is specific to the listed groups and endpoint. Broader conclusions require evidence beyond these short-term pharmacodynamic results.

15. Why This Trial Matters Statistically

This trial illustrates why a statistically strong laboratory result and a clinical conclusion are not interchangeable. The primary endpoint is a continuous pharmacodynamic measure, and the posted means show a clear descriptive separation between placebo and the PRT064445 regimens. The sponsor’s ANCOVA analyses also report small p-values. Those statistics address evidence for group differences under the model; they do not answer whether a laboratory change benefits patients.

The study also highlights the importance of distinguishing descriptive summaries from adjusted estimates. Group means and standard deviations help readers understand the observed values and within-group spread. ANCOVA estimates answer a model-based comparison that may account for covariates. Without the model specification and interval bounds, readers cannot fully assess the adjusted estimates’ precision or independently reproduce their interpretation.

Small active groups make careful attention to variability particularly important. The standard deviations differ across groups, and a mean alone cannot show the range of individual responses. Statistical methods need to be interpreted alongside sample size, variation, missing-data information, and the assumptions of the analysis. The registry does not report all of those details.

Multiplicity is another important issue in studies with several regimens and outcomes. Multiple comparisons can complicate interpretation if the testing strategy is not described. Here, the registry lists multiple analyses and secondary measures, but does not report a multiplicity procedure. That absence should prompt caution about treating individual p-values as a complete account of the evidence.

For biostatisticians and clinical researchers, the example reinforces a practical lesson: describe what the endpoint measures, distinguish the observed summaries from adjusted inference, and state the limits of generalization. The posted findings are informative about a short-term laboratory response in healthy volunteers, while leaving clinical effectiveness, durability, and broader safety questions unresolved.

16. 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 gives a concise description of the endpoint and sponsor analyses, but it does not report the complete statistical analysis plan. The ANCOVA covariates, model diagnostics, handling of unequal variability, and contrast coding are not reported. These omissions limit independent interpretation of the adjusted differences, especially their direction and the assumptions supporting the model.

Confidence-interval bounds are absent for the sponsor comparisons. Although the entries identify a 95% confidence level, without the interval limits readers cannot directly assess the precision of each adjusted estimate. The reported p-values alone do not convey the range of effects compatible with the model and data.

The reported active-regimen groups are small, and the standard deviations vary across groups. Small groups can yield imprecise estimates of variability and make results sensitive to individual observations. The record does not provide participant-level measurements, distributions, or sensitivity analyses that would allow readers to examine this concern.

The analysis population is described as subjects who received andexanet or placebo and were included in the pharmacodynamics analysis. Participant flow, missing endpoint data, exclusion reasons, and missing-data methods are not reported. The record also does not describe whether the results were analyzed according to a prespecified intention-to-treat or other population principle.

Multiplicity procedures are not reported despite the listed comparisons and secondary outcomes. The record does not provide the trial’s sample-size justification, the rationale for the enrollment target, or details of interim analyses. The sample-size planning assumptions are illustrations based on observed effects, not evidence of observed power or the original design rationale.

Finally, the endpoint is an early laboratory measure in healthy volunteers. It does not establish clinical outcomes, response duration, or applicability to patients. The record describes this analysis as Module 1 of a larger study, so conclusions should remain specific to the reported Module 1 endpoint and groups.

17. Related Tutorials

18. Related Statistical Calculators

19. Sources

• ClinicalTrials.gov: Phase 2 Healthy Volunteer Study to Evaluate the Ability of PRT064445 to Reverse the Effects of Several Blood Thinner Drugs on Laboratory Tests (Module 1 of 4) — NCT01758432.
• Registry record last updated 2023-08-07; status last verified 2023-08.
• Results first posted 2018-08-10.
• 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.

20. Record Summary

NCT01758432 is a completed phase 2 randomized, parallel-group study in healthy volunteers. The official title describes a masked, vehicle-controlled evaluation of intravenous PRT064445 after steady-state dosing with one of four direct or indirect fXa inhibitors. Module 1 intervention labels include apixaban and placebo.

The primary endpoint was percent change from baseline in anti-fXa activity at two minutes after the end of administration. The registry reports group means and standard deviations and sponsor ANCOVA comparisons of the listed PRT064445 regimens with placebo. Each reported comparison has p < 0.001; confidence-interval bounds are not reported.

The arm summaries show more negative mean changes in the PRT064445 groups than in placebo. These are short-term laboratory findings in healthy volunteers, not direct evidence about clinical outcomes. The record does not report multiplicity handling, the full ANCOVA specification, or missing-data methods.