Study of the Efficacy and Safety PF-06741086 in Adult and Teenage Participants With Severe Hemophilia A or Moderately Severe to Severe Hemophilia B: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Study of the Efficacy and Safety PF-06741086 in Adult and Teenage Participants With Severe Hemophilia A or Moderately Severe to Severe Hemophilia B (NCT03938792), a na phase 3 study. The primary outcome is Model-Based Annualized Bleeding Rate (ABR) of Treated Bleeding Events: Inhibitor Cohort (Participants With Prior OD Therapy at OP).

2. Clinical Question

3. Trial Design

NCT03938792 is a Phase 3, open-label crossover study sponsored by Pfizer. It enrolled 188 participants and is listed as completed. The record gives a start date of 2020-03-09 and a primary completion date of 2025-04-29. The study included people with hemophilia A or hemophilia B; its official title specifies adolescent and adult participants, with or without inhibitors.

The described sequence begins with a 6 month observation phase on participants’ current hemophilia regimen. Participants were then assigned to PF-06741086 treatment in the subsequent treatment period. The primary endpoint description distinguishes an observation-period reporting group, with a timeframe of up to 6 months, from an active-treatment-period reporting group, with a timeframe of up to 12 months. The registry does not report the actual treatment schedule or full sequence details.

“Open-label” means that treatment assignment was not masked. The crossover model indicates that the study compares periods within its treatment sequence, but the registry record does not give enough detail to reconstruct every participant’s sequence or the analysis of period and sequence effects. Allocation is listed as not applicable, and the record explicitly cautions that this is not a randomized comparison. Differences between periods should therefore not be interpreted as if they came from randomized, concurrent groups.

The primary endpoint concerns treated bleeds in a specific inhibitor subgroup with prior OD therapy during observation. Other listed outcomes include model-based annualized rates for joint, spontaneous, target-joint, and total bleeds in inhibitor and non-inhibitor cohorts, as well as changes in joint health and quality-of-life measures. The record lists 33 outcome measures and 40 analyses. Details of how all outcomes and analyses were prioritized 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: Data from participants with inhibitors who were on RP therapy were pooled for safety analyses only and were not included in efficacy hypothesis testing.

Observational Phase

Reasons for not completing this period:

Active Treatment Phase

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Modified Intent-to-Treat (mITT) set included all participants who completed OP and received at least 1 dose of PF-06741086 in ATP (excluding participants with inhibitors who were treated with routine prophylaxis in the OP). Participants who changed from a non-inhibitor to an inhibitor on or before ATP Day -7 testing were excluded from mITT. Here, "Overall Number of Participants Analyzed" signifies participants in mITT 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.

6. Primary Endpoint

The registry lists the primary outcome as Model-Based Annualized Bleeding Rate (ABR) of Treated Bleeding Events: Inhibitor Cohort (Participants With Prior OD Therapy at OP), measured over OP reporting group: Up to 6 months; ATP reporting group: up to 12 months.

The primary endpoint is the model-based annualized bleeding rate (ABR) of treated bleeding events in the inhibitor cohort with prior OD therapy during the observation phase. ABR expresses the number of bleeding episodes per year. For this endpoint, a bleed counted as treated if factor replacement or a bypass agent was used within 48 hours of bleeding onset, regardless of whether the treatment was preventive, prophylactic, or OD medication.

The registry defines the rate as the number of bleeds requiring treatment divided by days in the treatment period, multiplied by 365.25. If a participant did not complete a treatment period, the period ended at the last dosing date plus 6 days. Annualization makes bleeding frequency interpretable on a common yearly scale when participants contribute different amounts of follow-up. It does not mean that every participant was observed for a full year.

The endpoint focuses on treated events, not every reported bleed. That distinction matters because the outcome depends on both bleeding and the application of the registry’s treatment rule. The model-based estimate is reported with a 95% confidence interval. The registry record describes this endpoint and its model but does not report the estimate or interval, so the rate and its precision cannot be assessed from the record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does annualized bleeding rate mean?

ABR expresses the frequency of bleeding episodes on a yearly scale. In this study, the primary endpoint counts bleeds that met the registry’s treatment definition: factor replacement or a bypass agent within 48 hours of bleeding onset. The stated calculation uses treated bleeds divided by days in the treatment period, with annualization by 365.25. Annualization allows rates to be expressed on a common time scale even when observation time differs. It is a rate measure, not a count of events experienced by every participant over a full year.

Why use a negative binomial model?

The registry reports a repeated-measures negative binomial model for the primary outcome. In general, a negative binomial model is used for count outcomes when the variation in counts is greater than a simpler count model would accommodate. This is often called over-dispersion. Accounting for over-dispersion can make model-based uncertainty more appropriate for variable event counts. This general explanation does not establish which model checks were performed in this trial; those details are not reported in the registry record.

“Repeated measures” indicates that the model is intended to handle multiple observations or periods associated with participants. In general, measurements from the same participant may be related rather than independent. A repeated-measures approach can account for that dependence, but the registry record does not describe the model’s exact terms, correlation structure, covariates, or estimation procedure. Those details cannot be reconstructed from the model name alone.

How should a confidence interval be read?

The primary outcome is specified with a 95% confidence interval. In general, a confidence interval describes the statistical uncertainty around an estimate under the model and its assumptions. It helps readers assess how precisely an outcome has been estimated. It is not a guarantee that the true value lies within the interval, and it does not by itself show whether an effect is clinically important. The registry record does not provide the primary estimate or interval, so their width and position cannot be interpreted.

What does the crossover structure add?

The registry describes an observation phase on current treatment followed by PF-06741086 treatment, and labels the study model as crossover. In general, crossover designs can compare periods within participants, which may reduce the influence of stable differences between participants. They also require careful consideration of time, sequence, and possible carryover effects. The record does not report a washout period, sequence details, or how those potential effects were handled. No assumption about these design features should be made from the word “crossover” alone.

What is the analysis population?

The primary endpoint description identifies a modified intent-to-treat set. It includes participants who completed the observation phase and received at least 1 dose of PF-06741086 in the active treatment period, while excluding participants with inhibitors who received routine prophylaxis during observation. The record begins to describe additional rules for participants who changed inhibitor status, but that description is incomplete. The complete population definition and the number analyzed for the primary endpoint are not reported in the registry record.

The model-based result would describe the specified analysis population and endpoint, not automatically every enrolled participant. In general, analysis-population definitions matter because eligibility for an analysis can depend on treatment exposure, completion, or clinical subgroup. The registry’s incomplete population description limits assessment of precisely who contributed to the primary estimate.

What cannot be inferred from planning calculations?

Any sample-size illustrations based on an observed effect are planning examples, not observed trial findings and not observed power. The record says the illustration assumes a two-sided significance level of 0.05, equal allocation, and a normal approximation, with the effect to detect expressed as a fraction of the observed effect. These assumptions describe the illustration, not necessarily the trial’s design. The trial’s own sample-size justification is not reported in the registry record. No numerical planning result is stated in this explanation.

More generally, planning estimates depend on assumptions about the effect to detect and variability. They do not establish that a study was adequately powered for its actual result, nor do they replace a reported confidence interval. The registry record does not supply the primary endpoint estimate needed to interpret such an effect-based illustration in relation to the observed outcome.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists 33 outcome measures and 40 analyses, including several annualized bleeding rates, subgroup-specific joint-health and quality-of-life outcomes, and safety outcomes. When a study evaluates multiple outcomes or analyses, the chance of at least one apparently notable result can increase. In general, multiplicity methods can address that issue by adjusting inference or by defining a hierarchy of outcomes. Whether this study used such methods is not reported in the registry record. The primary endpoint’s status as primary is stated, but a complete testing hierarchy and multiplicity adjustment plan are not reported.

The registry gives a modified intent-to-treat description for the primary endpoint, including completion of the observation phase and receipt of at least 1 dose in the active treatment period, with an exclusion for inhibitor participants on routine prophylaxis during observation. The description of participants who changed inhibitor status is cut off. The complete analysis-population rules and the number included in each primary analysis are not reported in the registry record.

How missing outcome data were handled is also not reported. The endpoint description specifies how treatment-period days were ended for participants who did not complete a period, but that does not explain how missing bleeding information or other missing measures were treated in the model. The record does not state whether imputation, sensitivity analyses, or another missing-data procedure was used.

The registry does not report interim analyses, stopping rules, or the trial’s full statistical analysis plan. Their absence from the registry record should not be interpreted as evidence that a particular procedure was or was not used. It means those methods cannot be evaluated from the information recorded there.

11. 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 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: gastric haemorrhage (2.7%), rash (2.1%), vomiting (1.8%), device related infection (1.8%), traumatic haemorrhage (1.8%).

The adverse-event record separates participants by inhibitor status, treatment during observation, and the period or regimen shown in each group label. The entries report the number at risk for serious adverse events, the number affected by serious adverse events, the number at risk for other adverse events, the number affected by other adverse events, and deaths. These are distinct group-period entries; they should not be added together as if they represented independent groups.

For example, the inhibitor cohort’s OD-at-observation group receiving PF-06741086 in the active treatment period [OD at observation] lists 48 at risk for serious events, 1 affected by a serious event, 48 at risk for other events, 27 affected by other events, and 0 deaths. The non-inhibitor cohort’s routine-prophylaxis-at-observation group receiving PF-06741086 in the active treatment period lists 83 at risk for serious events, 7 affected by serious events, 83 at risk for other events, 31 affected by other events, and 0 deaths.

The record lists gastric haemorrhage at 2.7%, rash at 2.1%, vomiting at 1.8%, device related infection at 1.8%, traumatic haemorrhage at 1.8%, and haemarthrosis at 1.8% among its top serious-event terms. The denominator and event-term coding details for those percentages are not reported in the registry record. They should not be interpreted as comparative risks between treatments. The recorded deaths are 0 in each listed group, but these entries alone do not establish broader safety conclusions.

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

The registry record describes a completed Phase 3 study that examined PF-06741086 after an observation phase on participants’ current hemophilia regimens. It identifies a primary model-based annualized rate of treated bleeds in a specified inhibitor cohort and describes a repeated-measures negative binomial analysis. These details clarify what the study set out to measure and how the primary outcome was characterized.

The record does not report the primary numerical result, confidence interval, or p-value. It therefore does not establish the size, direction, or precision of the primary outcome. It also does not support a claim that the treatment reduced bleeding, increased bleeding, or had no effect. The sponsor’s summary of anticipated benefit is a rationale, not a report of observed results.

The study’s nonrandomized, open-label crossover structure is important when interpreting comparisons across periods. Participants received their current regimen during observation and PF-06741086 afterward, so the registry’s stated caution against treating the arms as like-for-like controls should be kept in view. The information recorded does not show how time-related or other potential differences were handled analytically.

The safety entries describe adverse events in the listed groups, including serious and other events, and show 0 deaths in each listed group. They are not a randomized comparison and should not be treated as proof of relative safety. The registry record also does not provide the context needed to interpret every listed event percentage as a comparative rate.

In short, the record permits a careful account of the trial’s population, sequence, endpoint definition, analysis model, and selected safety counts. It does not provide enough outcome information to quantify treatment effects or draw a conclusion about clinical benefit or harm.

13. Why This Trial Matters Statistically

This study is statistically notable for combining a crossover structure with a count-based bleeding endpoint. The primary outcome is an annualized rate of treated events, rather than a continuous measurement or a simple proportion of participants with an event. Its definition also ties event classification to treatment within 48 hours, making the operational rule central to what the endpoint represents.

The use of a repeated-measures negative binomial model reflects two statistical features relevant to this question: event counts can be more variable than a basic count model accommodates, and measurements across periods from the same participant may be related. These are general reasons such models are used; the registry record does not report the full model specification or diagnostic checks. The absence of the estimate and confidence interval means readers cannot assess how much uncertainty remained around the primary result.

The design also makes interpretation dependent on the comparison structure. A sequential observation and treatment framework may be informative about change across periods, but it is not a randomized, concurrent control comparison. The registry explicitly flags that limitation. Statistical interpretation should therefore distinguish a model-based outcome estimate from a causal effect that would require stronger design and analytic support.

Finally, the record lists multiple bleeding, joint-health, quality-of-life, and safety outcomes. That breadth makes transparent reporting of outcome priorities, multiplicity handling, analysis populations, and missing-data methods important. Several of those details are not reported in the registry record. The study illustrates why endpoint definitions and design labels are only part of a statistical assessment: the numerical results and analysis details are also needed to judge precision and interpretability.

14. 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 most consequential limitation of the registry record for statistical interpretation is that it does not report the primary endpoint estimate or its 95% confidence interval. Without those values, neither the magnitude nor the precision of the primary outcome can be evaluated. A completed status does not fill that reporting gap.

The study is open-label and nonrandomized, with participants moving from their current regimen during observation to PF-06741086 treatment. The registry cautions that the arms are not a like-for-like control comparison. Potential influences related to period, sequence, or changes over time cannot be assessed from the available design details. The registry record does not report a washout period or explain how such influences were addressed.

The primary endpoint applies to a defined inhibitor subgroup with prior OD therapy during observation, not automatically to all enrolled participants or all listed hemophilia cohorts. The modified intent-to-treat description is incomplete where it begins to address changes in inhibitor status. The number contributing to the primary analysis is not reported in the registry record.

The record does not report how missing outcome data were handled, whether multiplicity was controlled, or the complete statistical analysis plan. It also does not report the trial’s sample-size justification or interim-analysis procedures. These absences limit evaluation of the robustness and precision of the reported analyses.

Safety entries are divided into cohort and treatment-period groups, and the listed event percentages lack a reported denominator in this record. Because the groups are not a randomized comparison and may represent different periods or overlapping study populations, they should not be combined or read as direct comparative risks.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Study of the Efficacy and Safety PF-06741086 in Adult and Teenage Participants With Severe Hemophilia A or Moderately Severe to Severe Hemophilia B — NCT03938792.
• Registry record last updated 2026-08-14; status last verified 2026-07.
• Results first posted 2026-08-14.
• 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.

18. Record Summary

NCT03938792 is a completed Phase 3, open-label crossover study sponsored by Pfizer, with an enrollment of 188. It studied PF-06741086 in adolescent and adult participants with severe hemophilia A or moderately severe to severe hemophilia B, with or without inhibitors. Participants had a 6 month observation phase on their current regimen before PF-06741086 treatment.

The primary outcome was the model-based annualized rate of treated bleeding events in an inhibitor cohort with prior OD therapy during observation. The registry describes a repeated-measures negative binomial model and a 95% confidence interval, but does not report the primary estimate or interval. It lists additional bleeding, joint-health, quality-of-life, and safety outcomes.

The comparison was not randomized, and the registry cautions against treating the arms as like-for-like controls. The record provides selected adverse-event counts and terms, but key details on multiplicity, missing-data handling, and the complete analysis population are not reported.