Pivotal Study (Pharmacokinetics, Efficacy, Safety) of BAX 326 (rFIX) in Hemophilia B Patients: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Pivotal Study (Pharmacokinetics, Efficacy, Safety) of BAX 326 (rFIX) in Hemophilia B Patients (NCT01174446), a randomized phase 3 study. The primary outcome is Study Part 1- Area Under the Plasma Concentration Versus Time Curve From 0 to 72 Hours Per Dose.

2. Clinical Question

The clinical question is therefore focused on comparative pharmacokinetic exposure in the analyzed participants, rather than a direct comparison of clinical benefit. The study's broader purpose included hemostatic efficacy and safety, but the primary posted endpoint is a measure of drug concentration over time. Secondary outcomes include additional pharmacokinetic measures, quality-of-life scores, and pain assessment. Those outcomes provide context about what else was assessed, but they do not change the nature of the primary question.

3. Trial Design

The registry record describes a Phase 3 randomized, double-blind, crossover study of BAX 326 and BeneFIX, both recombinant factor IX products. It lists 86 participants as enrolled and identifies hemophilia B as the condition studied. Participants were previously treated and had severe or moderately severe disease. The record identifies Baxalta, now part of Shire, as the sponsor.

The crossover design means the study was organized to assess both products within a crossover framework rather than as a simple parallel-group comparison. The registry record does not report the treatment sequence, the length or use of a washout period, or the allocation ratio. It also does not describe how the order of product exposure was incorporated into the analysis. These details matter when interpreting crossover studies, because results can depend on how treatment periods and within-participant comparisons are handled.

The primary outcome was measured in Study Part 1 and defined for the pharmacokinetic per-protocol analysis population. That population included participants who took part in Study Parts 1–3, completed Study Part 1, and had no major protocol deviations. The analysis population reported for this endpoint was 25 participants for each product. This is not the same as the overall enrollment count: enrollment and endpoint-specific analysis counts describe different stages and populations.

The registry record lists study start as 2010-07-29, primary completion as 2012-05-03, and overall status as completed. It also lists 50 outcome measures and 32 analyses. The record does not report the planned sample-size justification, the treatment-period schedule, or whether interim analyses were conducted. These design details should not be inferred from the crossover label or from the observed numbers.

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: 86 enrolled. 1 withdrew consent prior to randomization. -31 were enrolled and randomized in Parts 1-3. Prior to receiving study drug: 3 discontinued by participant. -Part 2 an additional 54 were enrolled. Prior to receiving study drug: 3 discontinued by participant, 1 withdrew due to an AE, 2 screen failures, and 3 withdrew due to site closure.

Part 1 -Pharmacokinetic Crossover

Part 2 -BAX326 Prophylaxis and On-demand

Reasons for not completing this period:

Part 3 -Pharmacokinetic BAX326 Only

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Pharmacokinetic Per Protocol Analysis Set (PKPPAS) -Participants who participated in Study Parts 1-3 and completed Study Part 1 without any major protocol deviations

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.

6. Primary Endpoint

The registry lists the primary outcome as Study Part 1- Area Under the Plasma Concentration Versus Time Curve From 0 to 72 Hours Per Dose, measured over 72 hours.

The primary endpoint measures the area under the plasma concentration-versus-time curve from 0 to 72 hours per dose in Study Part 1. In general, an area-under-the-curve measure combines concentration and time: it represents drug exposure over the specified observation window rather than a concentration at a single sampling time. Expressing the measure per dose allows the reported exposure to be interpreted relative to the administered dose.

The registry record says the endpoint was computed using the linear trapezoidal method. The concentration at 72 hours was interpolated from nearby sampling times or extrapolated using the last quantifiable concentration and the terminal rate constant. That rate constant was estimated from a natural log-linear fit to later quantifiable concentrations, selecting the fit with the largest adjusted R-squared. These details describe how the endpoint was derived from concentration measurements; they do not, by themselves, describe the comparative statistical test.

Pharmacokinetic exposure is useful for characterizing how a product behaves in the body. It is not a direct measure of bleeding control, quality of life, or safety. The record lists hemostatic efficacy and safety among the study aims, but the primary endpoint is pharmacokinetic. The endpoint analysis population was the pharmacokinetic per-protocol set, not necessarily every enrolled participant.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The main comparative result is a ratio of geometric means. In general, a geometric mean is suited to measurements that are positive and may have a right-skewed distribution; it is calculated on a logarithmic scale and then expressed back on the original scale. A ratio of geometric means describes relative values, unlike a difference, which describes an absolute distance between values. The registry reports the ratio as 1.063 and its 90% confidence interval as 1.03 to 1.09. It does not report a p-value for this primary analysis or a detailed method used to construct the interval.

What does the endpoint measure?

The primary measure is area under the plasma concentration-versus-time curve from 0 to 72 hours per dose. In general, this integrates concentration across the observation period, so it reflects exposure over time rather than a single concentration. The registry specifies the linear trapezoidal method for the area calculation. It also describes how the concentration at 72 hours and the terminal rate constant were estimated when direct observations were not available at the endpoint of the window. These computational choices define the derived endpoint; they do not establish how the two products were compared statistically.

Why do the summary measure and comparative estimate differ?

The product-level primary summaries are medians with inter-quartile dispersion, whereas the comparative estimate is a ratio of geometric means. These are distinct summaries and should not be treated as if one were obtained by dividing the posted medians. A median describes the midpoint of the observed distribution. A geometric mean summarizes values after logarithmic transformation and is used in the reported ratio. Differences between these kinds of summaries are not inherently inconsistent.

The posted primary analysis is identified as a non-inferiority-or-equivalence legacy analysis, but no margin is reported. In general, a non-inferiority assessment requires a prespecified margin defining how much difference would be acceptable, while an equivalence assessment requires prespecified bounds on both sides. The registry record does not state those bounds or explain the hypothesis-testing procedure. Thus, the analysis label alone does not tell the reader what threshold was used or whether a formal criterion was met.

What does a crossover design change?

In general, a crossover design can allow each participant to receive more than one product, making within-participant comparisons possible and accounting for some stable differences between participants. Valid interpretation also depends on period effects, sequence, and possible carryover from an earlier treatment period. The registry record identifies randomization, double masking, and crossover, but does not report the sequence, washout details, or the model used to account for period or sequence. The analysis method for the primary comparison is not described beyond its reported ratio-of-geometric-means parameter.

How should recalculated comparisons be read?

Some secondary comparisons use simple unadjusted recalculations from the posted group summaries. Such calculations compare displayed summary statistics and do not reproduce a paired analysis unless the paired data or the necessary within-participant covariance is available. The sponsor-reported secondary analyses include paired t-tests. A paired test uses the within-participant differences and their variability; a comparison reconstructed from separate group means and standard deviations does not necessarily use that same information. Therefore, the resulting estimates and p-values may differ even when both use the same participants.

Unadjusted recalculations are descriptive checks, not replacements for the sponsor's analysis plan. They do not account for covariates, repeated observations, or a multiplicity strategy unless those elements are explicitly included. In this record, the recalculated results are not flagged as discrepant from the sponsor's own analysis. Where a sponsor analysis is reported, its method and result should be interpreted as the registry's reported analysis, while recalculated summaries remain a separate, simpler view.

Why does over-dispersion matter?

In general, over-dispersion occurs when observed variability is greater than a model assumes. If a statistical model understates variability, standard errors can be too small and confidence intervals can be too narrow, making evidence appear more precise than it is. The posted primary result does not report a dispersion-model assessment, and the record does not say whether over-dispersion was considered. This is a general statistical consideration, not a claim that over-dispersion occurred in this study.

How do planning illustrations differ from power?

In general, a sample-size calculation uses assumptions about the effect to detect, variability, significance level, and allocation to estimate a planning target. The posted planning illustrations are based on observed effects and specified assumptions, including a two-sided significance level of 0.05, equal allocation, and a normal approximation. They are illustrations rather than the trial's design justification and are not observed power. The registry record does not report the study's original sample-size justification. Sample-size planning values should not be used to infer the probability that the trial detected an effect after observing its results.

10. Analyses Reported by the Sponsor

The registry reports a primary ratio of geometric means of 1.063, with a 90% confidence interval from 1.03 to 1.09. It identifies the analysis as a legacy non-inferiority-or-equivalence type, but does not provide a p-value, margin, or full decision rule. The sponsor's reported primary analysis is therefore the central comparative result available in the record, while the formal criterion behind its interpretation cannot be reconstructed from the posted information.

The recalculated secondary comparisons use unadjusted summaries, whereas some sponsor-reported secondary results use paired t-tests. These methods do not necessarily estimate the same quantity: paired analyses use within-participant changes, while a simple recalculation from separate group summaries does not necessarily preserve that pairing. The registry does not flag a discrepancy between recalculated and sponsor analyses. That absence of a flag does not make the methods interchangeable; it means the reported data do not identify a specific conflict requiring a correction to the sponsor result.

For the primary outcome, no separate recalculation is reported. The ratio and confidence interval should be read as the sponsor-reported comparison, not as an independent reconstruction. Any formal claim about non-inferiority or equivalence would require the margin and decision rules, which are not reported in the registry record.

11. Multiplicity, Missing Data and Analysis Populations

The registry record lists 50 outcome measures and 32 analyses, with 49 secondary outcomes posted. It does not report whether the study used a multiplicity adjustment, a hierarchy of hypotheses, or another prespecified strategy for controlling the chance of false-positive findings across multiple analyses. In general, when many outcomes are tested separately, some nominally small p-values can arise by chance even if there is no underlying effect. For that reason, secondary results without an adjustment or prespecified hierarchy should be treated as less confirmatory than a single primary analysis.

The record does not report a full missing-data plan, including how missing outcome measurements were handled, whether any imputation was used, or how participants who did not complete all study parts were treated in each analysis. The primary population is explicitly restricted to participants who participated in Study Parts 1–3, completed Study Part 1, and had no major protocol deviations. That definition is useful, but it does not explain the reasons that other enrolled participants were not included or how missing values within the analysis population were handled.

Interim analyses are not reported in the registry record. Nor is a complete statistical analysis plan reported. The absence of those details should not be interpreted as evidence that such procedures did or did not occur; the record simply does not state them.

These omissions limit assessment of how analysis choices may affect precision and the risk of selective interpretation. In particular, readers cannot verify whether the endpoint hierarchy, missing-data rules, and multiplicity approach were established before results were known. The reported values remain useful descriptions of the posted analyses, but the registry record does not provide enough detail to independently audit all aspects of the statistical plan.

12. 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: intestinal obstruction (1.4%), cervical vertebral fracture (1.4%), traumatic haematoma (1.4%), hepatitis b core antibody positive (1.4%), convulsion (1.4%).

The posted safety figures list 73 participants at risk for serious adverse events in the BAX326 group and 4 participants affected. For other adverse events, 73 were at risk and 15 were affected. The record also lists five serious adverse event terms, each reported at 1.4%: intestinal obstruction, cervical vertebral fracture, traumatic haematoma, hepatitis B core antibody positive, and convulsion.

These counts describe events recorded in the registry; they do not establish that an event was caused by the product. The listed terms are event categories, and the record does not provide enough detail to assess timing, severity beyond the serious-event classification, recurrence, or attribution. The death count is not reported in the registry record.

The safety group shown is BAX326, and the posted figures do not provide a corresponding comparator-group summary in these safety data. They therefore cannot support a direct comparative safety conclusion. Also, safety counts have a different denominator and purpose from the primary pharmacokinetic analysis. They should be read as descriptive registry reporting rather than as a controlled estimate of relative risk.

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

The primary result describes pharmacokinetic exposure over a specified 72-hour interval in the per-protocol analysis population. It provides a ratio of geometric means and a 90% confidence interval, which communicate the estimated relative exposure and the uncertainty around that estimate. The result does not directly measure hemostatic efficacy, bleeding outcomes, quality of life, or safety.

The study is identified as randomized, double-blind, and crossover, but the registry record does not report all details needed to reconstruct the analysis or its formal decision criteria. In particular, a non-inferiority or equivalence margin is not reported. The posted estimate should therefore not be taken alone as proof that a formal non-inferiority or equivalence criterion was met. The confidence interval is informative about the ratio under the reported analysis, but it cannot supply an unreported margin.

The endpoint population is narrower than the overall enrolled population. Its result pertains to participants who met the specified pharmacokinetic per-protocol criteria, and should not be generalized automatically to every enrolled participant or to populations not described in the registry. The secondary outcomes offer additional information, but multiple unadjusted analyses and differing methods reduce the strength of any isolated secondary signal.

More generally, a statistical estimate is not a treatment recommendation. The figures in the registry describe measured outcomes and the uncertainty attached to particular analyses. They do not determine what treatment is appropriate for an individual, and the registry record alone does not provide all clinical context needed to make such a judgment. The defensible interpretation is limited to the posted pharmacokinetic comparison and the descriptive secondary and safety information, with uncertainty about unreported design and analysis details kept explicit.

14. Why This Trial Matters Statistically

This study illustrates why a clinical trial result must be interpreted in relation to the endpoint and analysis population, not only the overall enrollment. The primary result concerns area under the plasma concentration-versus-time curve per dose, a pharmacokinetic measure. Its analysis used a defined per-protocol set, so its target population is narrower than the full enrolled group. Keeping those distinctions clear prevents a pharmacokinetic estimate from being mistaken for a direct clinical outcome.

The crossover design also highlights the importance of knowing how treatment periods and within-participant information were handled. A crossover can support efficient comparisons, but interpretation depends on period, sequence, and carryover considerations. The registry identifies the design but does not report enough detail to evaluate those elements in the primary model. This makes transparency about the full analysis plan particularly valuable.

The record also shows the importance of separating the sponsor's reported analysis from simpler unadjusted recalculations. Different statistical methods can address different comparisons, especially when paired data are involved. A recalculation from group summaries may be useful for orientation, but it cannot automatically reproduce a paired analysis or substitute for its assumptions.

Finally, the extensive set of outcomes emphasizes the need to distinguish confirmatory from supportive findings. Without a reported multiplicity strategy, secondary results should be interpreted cautiously. Together, these issues make this trial useful for understanding how endpoint choice, analysis populations, design structure, and reporting completeness shape the conclusions that can reasonably be drawn from a registry record.

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 does not report the non-inferiority or equivalence margin underlying the primary analysis, nor a complete decision rule. Although the comparative result is labeled as a legacy non-inferiority-or-equivalence analysis, that label is not enough to assess whether a formal criterion was satisfied. The primary analysis method is also not described in detail beyond the reported ratio of geometric means.

Important crossover details are not reported, including treatment sequence, washout arrangements, and how period or sequence effects were handled. These omissions limit independent assessment of whether the analysis fully reflects the design. The record also does not report the original sample-size justification, interim-analysis procedures, or a complete statistical analysis plan.

The primary endpoint uses a pharmacokinetic per-protocol population, with 25 participants reported for each product, rather than the full enrollment of 86. The record defines the per-protocol criteria but does not explain all exclusions or the handling of missing measurements. These limits affect the reader's ability to assess the representativeness and robustness of the endpoint analysis.

Secondary findings arise from numerous outcome measures and analyses. A multiplicity strategy is not reported, and the results include different analysis approaches, including paired tests and unadjusted recalculations. This makes isolated secondary results less suitable for confirmatory interpretation. Safety reporting is also limited in the posted figures: the described event counts are for the BAX326 group, and comparative safety information is not available in those figures. These limitations arise from the information in the registry record and should not be filled by assumptions.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Pivotal Study (Pharmacokinetics, Efficacy, Safety) of BAX 326 (rFIX) in Hemophilia B Patients — NCT01174446.
• Registry record last updated 2021-05-20; status last verified 2021-04.
• Results first posted 2013-11-25.
• 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

NCT01174446 is a completed Phase 3 randomized, double-blind crossover study of BAX 326 and BeneFIX in previously treated patients with severe or moderately severe hemophilia B. The registry record lists 86 participants enrolled and identifies pharmacokinetics, efficacy, safety, and immunogenicity among the study aims.

The primary endpoint was area under the plasma concentration-versus-time curve from 0 to 72 hours per dose in Study Part 1. The posted primary comparison reports a ratio of geometric means of 1.063, with a 90% confidence interval from 1.03 to 1.09, in the pharmacokinetic per-protocol population. The registry does not report the margin or complete decision rule needed to independently assess a formal non-inferiority or equivalence conclusion.

The record also includes secondary outcomes and descriptive safety figures. Multiplicity handling, the full missing-data plan, interim analyses, and the complete statistical analysis plan are not reported in the registry record.