This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. Numerical trial results on this page are restricted to the results posted in the ClinicalTrials.gov record; calculations added by Clinical Biostats are labelled as such.
1. Trial at a Glance
This page covers the registry results for Safety and Tolerability of Abelacimab (MAA868) vs. Rivaroxaban in Patients With Atrial Fibrillation (NCT04755283), a randomized phase 2 study. The primary outcome is Incidence Rate of the First Occurrence of Composite of International Society on Thrombosis and Haemostasis (ISTH)-Defined Major Bleeding or Clinically Relevant Non-Major (CRNM) Bleeding Events.
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT04755283 |
| Phase | Phase 2 |
| Condition | Atrial Fibrillation (AF), Stroke |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | Triple |
| Enrollment | 1287 |
| Arms in the primary outcome | 3 |
| Lead sponsor | Anthos Therapeutics, Inc. |
| Results posted | Yes |
2. Clinical Question
Population
Patients with atrial fibrillation at moderate-to-high risk of stroke were the target population. The primary analysis included randomized participants who received at least one dose of study treatment. The registry lists atrial fibrillation and stroke as conditions. It does not report further eligibility details in the primary outcome description.
Intervention
Abelacimab was studied in two dose groups. The 90 mg group received low-dose subcutaneous treatment monthly during the randomized phase, with high-dose monthly treatment described for the extension. The 150 mg group received high-dose subcutaneous treatment monthly, including during the extension.
Comparator
The comparator was rivaroxaban, described as 20 mg orally once per day with the evening meal. The record describes dose adaptation to 15 mg daily for patients with a Creatinine Clearance of 50 ml/min or lower by the Cockcroft-Gault equation.
Primary question
How did the incidence rate of a first ISTH-defined major or clinically relevant non-major bleeding event compare for each abelacimab dose versus rivaroxaban? The endpoint expresses first events relative to person-time during the randomized phase, rather than simply counting all events.
The clinical question is comparative and focused on bleeding. The registry summary says the study was intended to evaluate the bleeding profile of abelacimab relative to rivaroxaban in people with atrial fibrillation at moderate-to-high risk of stroke. The primary endpoint combines major bleeding and clinically relevant non-major bleeding as defined by the International Society on Thrombosis and Haemostasis. The posted secondary outcomes separately address major bleeding and a broader combination of major, CRNM, or minor bleeding.
The comparison has three arms: abelacimab 90 mg, abelacimab 150 mg, and rivaroxaban. The sponsor-reported primary analyses compare each abelacimab group with rivaroxaban. They do not report a direct comparison between the two abelacimab doses. The study record does not report a primary question about stroke prevention or establish a conclusion about stroke outcomes.
The trial was randomized and parallel-group. Its phase is listed as Phase 2. The registry describes triple masking, while the title identifies rivaroxaban as open-label; the record does not explain how those masking descriptions relate to one another. The question should therefore be read in the context of the posted design and the specific bleeding endpoint, rather than as a general comparison of all potential benefits and harms.
3. Trial Design
AZALEA is a randomized, active-controlled, parallel-group Phase 2 study sponsored by Anthos Therapeutics, Inc. The registry lists enrollment of 1287 and the status as active, not recruiting. The recorded start date is 2021-02-02, and the primary completion date is 2024-02-15. The primary endpoint description states that assessment occurred at visits from randomization through the end of the randomized phase, up to 33 months.
The three assigned groups were abelacimab 90 mg, abelacimab 150 mg, and rivaroxaban. Abelacimab was administered subcutaneously each month. For the 90 mg group, the randomized-phase description calls the dose low; the extension description calls for high-dose monthly treatment. The 150 mg group received high-dose monthly treatment in both descriptions. Rivaroxaban was given orally once per day with the evening meal, with a stated dose adaptation for patients whose Creatinine Clearance was 50 ml/min or lower by the Cockcroft-Gault equation.
The registry identifies the study as randomized and triple-masked, and its model as parallel. The official title also describes rivaroxaban as open-label. These descriptions are both part of the record; further details about masking procedures are not reported in the registry record. The trial’s allocation concealment procedures and the specific roles of masking are likewise not reported in the registry record.
The primary analysis population is defined as randomized participants who received at least one dose of study treatment. The registry notes that the difference between the overall number analyzed and the overall number assigned represents those who never received a dose, but the sentence describing this population is truncated. The primary outcome lists 425 participants in the abelacimab 90 mg group, 427 in the 150 mg group, and 428 in the rivaroxaban group. These are outcome-specific analyzed group counts and should not be treated as a substitute for a complete participant-flow account.
The registry record does not report the protocol’s sample-size justification, planned recruitment assumptions, or detailed analysis plan. The posted sponsor comparisons use a stratified log-rank test and report hazard ratios for each abelacimab dose versus rivaroxaban.
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.
Overall Study
| Stage | Abelacimab 90 mg (MAA868) | Abelacimab 150 mg (MAA868) | Rivaroxaban |
|---|---|---|---|
| STARTED | 427 | 430 | 430 |
| Entered Open-Label | 250 | 240 | 215 |
| COMPLETED | 382 | 391 | 389 |
| NOT COMPLETED | 45 | 39 | 41 |
Reasons for not completing this period:
| Reason | Abelacimab 90 mg (MAA868) | Abelacimab 150 mg (MAA868) | Rivaroxaban |
|---|---|---|---|
| Withdrawal by Subject | 14 | 3 | 6 |
| Lost to Follow-up | 0 | 0 | 1 |
| Death | 31 | 36 | 34 |
Analysis population for the primary outcome, as defined in the registry: The analysis population is the number of randomized participants who received at least one dose of study treatment. The difference between the overall number of participants analyzed and the overall number of participants assigned to the arm represents those who never received a dose of study treatment.
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.
| Characteristic | Unit | Abelacimab 90 mg (MAA868) | Abelacimab 150 mg (MAA868) | Rivaroxaban | Total |
|---|---|---|---|---|---|
| Participants analysed | 427 | 430 | 430 | 1287 | |
| Age, Customized: ≥ 80 years | Participants | 87 | 81 | 90 | 258 |
| Sex/Gender, Customized: Female sex | Participants | 195 | 193 | 184 | 572 |
| Race/Ethnicity, Customized: White | Participants | 404 | 410 | 404 | 1218 |
| Race/Ethnicity, Customized: Asian | Participants | 20 | 15 | 24 | 59 |
| Race/Ethnicity, Customized: Black | Participants | 3 | 5 | 2 | 10 |
| Median Age | years | 75 (69 to 79) | 74 (69 to 78) | 74 (69 to 79) | 74 (69 to 78) |
| Region of Enrollment: North America | Participants | 126 | 106 | 103 | 335 |
| Region of Enrollment: Europe | Participants | 282 | 309 | 305 | 896 |
| Region of Enrollment: Asia | Participants | 19 | 15 | 22 | 56 |
| Median body-mass index (IQR) | kg/m² | 29.5 (26.0 to 33.6) | 30.0 (26.7 to 34.5) | 30.3 (27.0 to 34.4) | 29.9 (26.6 to 34.2) |
| Creatinine clearance ≤50 ml/min | Participants | 86 | 90 | 88 | 264 |
| Pattern of atrial fibrillation: First detected or paroxysmal | Participants | 224 | 220 | 225 | 669 |
| Pattern of atrial fibrillation: Persistent or long-standing persistent | Participants | 87 | 84 | 97 | 268 |
| Pattern of atrial fibrillation: Permanent | Participants | 115 | 120 | 106 | 341 |
| Median CHA₂DS₂-VASc score | score | 5.0 (4.0 to 5.0) | 5.0 (4.0 to 5.0) | 5.0 (4.0 to 6.0) | 5.0 (4.0 to 5.0) |
| CHA₂DS₂-VASc score: ≤4 | Participants | 191 | 213 | 186 | 590 |
| CHA₂DS₂-VASc score: 5 | Participants | 135 | 131 | 130 | 396 |
| CHA₂DS₂-VASc score: ≥6 | Participants | 101 | 86 | 114 | 301 |
| Median HAS-BLED score | score | 2.0 (2.0 to 3.0) | 2.0 (2.0 to 3.0) | 3.0 (2.0 to 3.0) | 3.0 (2.0 to 3.0) |
| HAS-BLED score ≥3 | Participants | 210 | 213 | 225 | 648 |
| Hypertension | Participants | 410 | 417 | 418 | 1245 |
| Diabetes | Participants | 223 | 231 | 245 | 699 |
6. Primary Endpoint
The registry lists the primary outcome as Incidence Rate of the First Occurrence of Composite of International Society on Thrombosis and Haemostasis (ISTH)-Defined Major Bleeding or Clinically Relevant Non-Major (CRNM) Bleeding Events, measured over Assessed at every visit from randomization through the end of the randomized phase of the study, up to 33 months..
| Element | Registry entry |
|---|---|
| Result type | Number |
| Measure of spread | not applicable |
| Unit | participants with event/100 person-years |
| Endpoint classification | not reduced to a single comparison |
The primary endpoint is the incidence rate of the first occurrence of an ISTH-defined major bleeding event or a clinically relevant non-major bleeding event. It combines two categories of bleeding events into one outcome. The measure is stated as participants with an event per 100 person-years. The registry describes the numerator as participants experiencing a first qualifying event and the denominator as person-years through the first event or the end of treatment across participants. For participants without an event, person-time runs through the end of treatment in the randomized phase.
Using person-time makes the rate sensitive to both the number of first events and the time participants were observed. This can be useful when participants have different lengths of follow-up: a simple proportion does not represent time under observation in the same way. The endpoint is therefore not just a count of participants with bleeding, nor is it a count of every bleeding event. It concerns the first event meeting the composite definition.
In general, a composite endpoint can capture more than one clinically defined event category, but its overall result does not by itself show how much each component contributed. The registry posts secondary outcomes for first ISTH-defined major bleeding and for first major, CRNM, or minor bleeding. Those outcomes provide distinct endpoint definitions; their results are not included in the primary result details.
The primary endpoint was assessed at visits from randomization through the end of the randomized phase, up to 33 months. The registry does not report further adjudication procedures or details about event ascertainment in the primary outcome description. The posted incidence rates do not include confidence intervals in the primary outcome data, so their precision cannot be read directly from those rate values alone. The comparative hazard-ratio analyses have separately reported confidence intervals.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The posted primary incidence rates were 2.64 participants with a first qualifying event per 100 person-years for abelacimab 90 mg, 3.22 for abelacimab 150 mg, and 8.39 for rivaroxaban. These are rates for the composite of first ISTH-defined major or CRNM bleeding events in the primary analysis population. They describe the observed event rate relative to person-time, not the chance that an individual participant will experience an event over a fixed duration.
For the abelacimab 90 mg versus rivaroxaban comparison, the sponsor reported a hazard ratio of 0.314, with a 95% confidence interval from 0.192 to 0.513 and p < 0.001. For abelacimab 150 mg versus rivaroxaban, the reported hazard ratio was 0.382, with a 95% confidence interval from 0.243 to 0.602 and p < 0.001. Both analyses are identified as superiority comparisons and use a stratified log-rank test.
A hazard ratio compares the relative event hazard over follow-up under the analysis model; it is not the same quantity as a ratio of the posted incidence rates. A value below 1 indicates a lower estimated hazard in the first-named group relative to the comparator in these reported analyses. The confidence intervals express uncertainty around the estimated hazard ratio under the analysis. Both intervals lie below 1, and both reported p-values are below 0.001. Those features describe the sponsor’s inferential results; they do not establish the size of an individual participant’s absolute risk over a particular time.
The rate values and hazard ratios answer related but different descriptive and comparative questions. Rates provide an event frequency relative to person-time for each group. The hazard-ratio analyses compare event timing between groups under a stratified log-rank approach. They should not be conflated or used to generate an unreported absolute difference, risk ratio, or number needed to treat.
The record reports two sponsor comparisons, each against rivaroxaban. It does not report a direct statistical comparison between the abelacimab doses. Nor does the primary result alone give the component-specific rates for major bleeding and CRNM bleeding. Although two secondary outcomes are listed, their numerical results are not included in the registry record. The primary finding should therefore be interpreted narrowly: the registry reports lower observed composite event rates in both abelacimab groups than in the rivaroxaban group, alongside sponsor-reported hazard-ratio analyses favoring each abelacimab group for this endpoint.
The reported results do not by themselves establish an overall safety profile or a conclusion about outcomes beyond the defined bleeding endpoint. The study’s design, endpoint, follow-up, and analysis population all delimit what the comparison describes. A statistical result for this endpoint is not a general conclusion about all possible clinical outcomes.
9. Statistical Methods Explained
What does an incidence rate represent?
In general, an incidence rate relates event occurrence to time at risk. Here, the registry defines the measure as participants with a first qualifying event per 100 person-years. A participant contributes observation time until a first event or the end of treatment during the randomized phase; participants without events contribute time through the end of treatment. This approach reflects the amount of follow-up contributing to the rate, rather than treating all participants as if they had identical observation periods.
The endpoint counts first occurrences of the composite event, not every event that might happen after a first qualifying bleed. The numerator is therefore framed around participants experiencing a first event. The denominator is person-time. These choices matter: a rate is not a simple percentage, and its numerical value should not be described as the proportion of participants who bled.
What does a hazard ratio describe?
In general, a hazard is an event rate at a point in follow-up among those who have not yet had the event. A hazard ratio compares hazards between groups over the period represented by the analysis. The reported estimates are 0.314 for abelacimab 90 mg versus rivaroxaban and 0.382 for abelacimab 150 mg versus rivaroxaban. Because these are below 1, the estimated hazard was lower in the abelacimab group than in the comparator group in each reported analysis.
A hazard ratio is not a direct estimate of the difference in incidence rates, and it does not state how many participants would avoid an event. It is also not an absolute risk at a particular follow-up time. Interpretation depends on the time-to-event analysis and its assumptions. The registry names a stratified log-rank test but does not report the strata or further model details. Those details are not reported in the registry record.
How should the confidence intervals be read?
In general, a confidence interval shows the range of estimates compatible with the observed data under the method used, together with its stated coverage procedure. It conveys uncertainty and precision, not a guarantee that the true value lies within the interval. The 95% confidence interval for the 90 mg comparison is 0.192 to 0.513; for the 150 mg comparison, it is 0.243 to 0.602. Both reported intervals exclude 1, the reference value for equal hazards. The intervals also show that the point estimates are not exact values known without uncertainty.
The confidence intervals are for the hazard ratios, not for the incidence rates. The registry does not post interval estimates for the primary rates in the registry record. Consequently, it is not appropriate to infer rate precision from the hazard-ratio intervals or to treat the displayed rate estimates as exact.
What does the p-value add?
In general, a p-value measures how unusual data at least as extreme as the observed data would be under a specified null hypothesis and analysis procedure. It is not the probability that the null hypothesis is true, and it does not quantify clinical importance. The sponsor reports p < 0.001 for each comparison. This is evidence against the relevant equal-hazard null under the stated analysis, but it does not answer questions about absolute benefit, other outcomes, or the probability that a result will replicate.
The analyses are labeled superiority comparisons. That label identifies the direction and purpose of the reported testing; it does not eliminate uncertainty or replace consideration of the endpoint definition, design, and multiplicity of testing. The registry does not report a multiplicity adjustment or a complete testing hierarchy.
Why can over-dispersion matter?
In general, count outcomes can vary more across observations than a simple count model expects. This is called over-dispersion. When present, failing to account for it can make standard errors too small and confidence intervals too narrow. However, the primary outcome is analyzed as a time-to-first-event endpoint with a stratified log-rank test, and the registry does not report a count-model analysis or an over-dispersion assessment. Over-dispersion should not be assumed to have been present or to have affected these results.
More generally, the statistical approach should match how the outcome is measured. The person-time rate summarizes first events relative to follow-up, while the stratified log-rank analysis compares time to first event. Neither measure alone describes every feature of the bleeding experience. The record does not report further information on censoring, missing endpoint data, or the details of stratification.
10. Analyses Reported by the Sponsor
| Method | Estimate | 95% CI | P-value |
|---|---|---|---|
| Stratified Log Rank Test | Hazard Ratio (HR) 0.314 | 0.192 to 0.513 | < 0.001 |
| Stratified Log Rank Test | Hazard Ratio (HR) 0.382 | 0.243 to 0.602 | < 0.001 |
The sponsor’s reported primary analyses compare each abelacimab dose with rivaroxaban using a stratified log-rank test. For abelacimab 90 mg versus rivaroxaban, the reported hazard ratio is 0.314, with a 95% confidence interval of 0.192 to 0.513 and p < 0.001. For abelacimab 150 mg versus rivaroxaban, it is 0.382, with a 95% confidence interval of 0.243 to 0.602 and p < 0.001. The registry labels both comparisons as superiority analyses.
The registry also posts primary incidence rates for each group. These rates are descriptive summaries based on first events and person-time, whereas the sponsor’s hazard ratios come from time-to-event comparisons. The two forms of result are related but are not interchangeable. The hazard ratios should not be recalculated by dividing the incidence rates, and the rate values do not replace the sponsor’s analysis.
The record does not identify a discrepancy between a recalculation and the sponsor’s result. No recalculated hazard-ratio values are reported in the record details. Accordingly, the sponsor-reported analysis is the available inferential result. It should be interpreted as a comparison of the primary composite endpoint, not as proof of an effect on every bleeding category or on other clinical outcomes.
11. Multiplicity, Missing Data and Analysis Populations
The registry lists two secondary outcomes: first ISTH-defined major bleeding and first ISTH-defined major, CRNM, or minor bleeding. It also reports two primary comparisons, each contrasting an abelacimab dose with rivaroxaban. Considering multiple outcomes or comparisons can increase the chance of at least one apparently positive result when each is tested separately. In general, multiplicity procedures can address that issue through planned testing sequences or adjusted significance thresholds.
The registry record does not report whether the two primary comparisons were adjusted for multiplicity, whether a testing hierarchy was prespecified, or how the secondary outcomes were handled inferentially. It also does not report numerical results for the secondary outcomes in the registry record. The p-values for the primary comparisons should therefore be read as reported, without assuming a particular family-wise error procedure or a confirmatory testing hierarchy.
The primary analysis population is described as randomized participants who received at least one dose of treatment. This is a treated, randomized population definition, rather than a statement that all randomized participants were included regardless of treatment receipt. The record notes that participants assigned but never dosed account for the difference between overall assigned and analyzed numbers, but the population description is truncated. The full participant flow and the reasons for not receiving treatment are not reported in the registry record.
Details about missing primary endpoint observations, censoring rules, withdrawals, and sensitivity analyses are not reported in the registry record. These details matter in time-to-event studies because observed follow-up may end before an event occurs, and the handling of incomplete follow-up affects interpretation. The record also does not report an interim analysis plan, a stopping rule, or a sample-size justification.
Planning illustrations based on an observed effect are not observed power, and they do not establish the original design assumptions. The record’s stated planning assumptions describe 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. Those assumptions alone do not provide a protocol-based sample-size rationale or resolve multiplicity and missing-data questions.
12. Safety Results
The registry reports serious adverse events as affected participants divided by participants at risk.
| Arm | Participants with serious adverse events | Participants at risk | Percentage |
|---|---|---|---|
| Abelacimab 90 mg (MAA868) | 158 | 425 | 37.2% |
| Abelacimab 150 mg (MAA868) | 157 | 427 | 36.8% |
| Rivaroxaban | 167 | 428 | 39.0% |
Deaths
| Arm | Deaths (participants affected of at risk) |
|---|---|
| Abelacimab 90 mg (MAA868) | 31 of 427 |
| Abelacimab 150 mg (MAA868) | 36 of 430 |
| Rivaroxaban | 34 of 430 |
Other (non-serious) adverse events
Participants with at least one non-serious event at or above the registry's 5% reporting threshold.
| Arm | Participants affected of at risk |
|---|---|
| Abelacimab 90 mg (MAA868) | 215 of 425 |
| Abelacimab 150 mg (MAA868) | 257 of 427 |
| Rivaroxaban | 227 of 428 |
Most frequent serious adverse event terms
| Event term | Organ system | Abelacimab 90 mg (MAA868) | Abelacimab 150 mg (MAA868) | Rivaroxaban |
|---|---|---|---|---|
| Cardiac failure | Cardiac disorders | 26 of 425 | 20 of 427 | 18 of 428 |
| Atrial fibrillation | Cardiac disorders | 18 of 425 | 14 of 427 | 10 of 428 |
| Pneumonia | Infections and infestations | 10 of 425 | 10 of 427 | 13 of 428 |
| Death | General disorders | 9 of 425 | 4 of 427 | 6 of 428 |
| Chronic obstructive pulmonary disease | Respiratory, thoracic and mediastinal disorders | 10 of 425 | 2 of 427 | 6 of 428 |
| Acute myocardial infarction | Cardiac disorders | 6 of 425 | 3 of 427 | 7 of 428 |
| Gastrointestinal haemorrhage | Gastrointestinal disorders | 1 of 425 | 3 of 427 | 12 of 428 |
| Anaemia | Blood and lymphatic system disorders | 1 of 425 | 3 of 427 | 11 of 428 |
| Acute respiratory failure | Respiratory, thoracic and mediastinal disorders | 7 of 425 | 3 of 427 | 5 of 428 |
| Cardiac failure congestive | Cardiac disorders | 2 of 425 | 6 of 427 | 6 of 428 |
Most frequent non-serious adverse event terms
| Event term | Organ system | Abelacimab 90 mg (MAA868) | Abelacimab 150 mg (MAA868) | Rivaroxaban |
|---|---|---|---|---|
| COVID-19 | Infections and infestations | 41 of 425 | 48 of 427 | 45 of 428 |
| Atrial fibrillation | Cardiac disorders | 28 of 425 | 27 of 427 | 24 of 428 |
| Arthralgia | Musculoskeletal and connective tissue disorders | 23 of 425 | 26 of 427 | 27 of 428 |
| Hypertension | Vascular disorders | 24 of 425 | 26 of 427 | 20 of 428 |
| Back pain | Musculoskeletal and connective tissue disorders | 20 of 425 | 26 of 427 | 22 of 428 |
| Urinary tract infection | Infections and infestations | 27 of 425 | 24 of 427 | 12 of 428 |
| Anaemia | Blood and lymphatic system disorders | 18 of 425 | 17 of 427 | 24 of 428 |
| Contusion | Injury, poisoning and procedural complications | 10 of 425 | 27 of 427 | 15 of 428 |
| Headache | Nervous system disorders | 13 of 425 | 27 of 427 | 11 of 428 |
| Epistaxis | Respiratory, thoracic and mediastinal disorders | 11 of 425 | 9 of 427 | 27 of 428 |
Serious events with the highest arm-level rate: cardiac failure (6.1%), atrial fibrillation (4.2%), pneumonia (3.0%), gastrointestinal haemorrhage (2.8%), anaemia (2.6%).
The adverse-event record reports serious adverse events in 158 participants in the abelacimab 90 mg group, 157 in the 150 mg group, and 167 in the rivaroxaban group. The corresponding groups at risk are listed as 425, 427, and 428. Other adverse events are reported for 215, 257, and 227 participants, respectively, with the same group-specific at-risk counts. Deaths are reported as 31, 36, and 34, respectively.
These figures are counts of participants with the recorded categories, not estimates of a comparative treatment effect. They do not by themselves show event timing, duration, severity beyond the serious classification, or whether an event was considered related to study treatment. The registry does not report those details in the adverse-event details. Serious adverse events should also not be conflated with the primary bleeding endpoint: the primary endpoint has a specific definition and includes first major or CRNM bleeding events.
The record lists the most frequent serious-event terms as cardiac failure, 6.1%; atrial fibrillation, 4.2%; pneumonia, 3.0%; gastrointestinal haemorrhage, 2.8%; anaemia, 2.6%; and chronic obstructive pulmonary disease, 2.4%. These terms are not presented with group-specific values in the record details. They therefore do not support a comparison of those events between treatment groups. Counts and percentages are descriptive safety information; interpretation requires the definitions, follow-up, attribution, and group-level detail that are not reported in the registry record.
13. What the Result Does, and Does Not, Mean
The posted results describe a randomized comparison of two abelacimab doses with rivaroxaban for a specific bleeding endpoint in patients with atrial fibrillation. In the primary analysis, the incidence rates for first major or CRNM bleeding were 2.64, 3.22, and 8.39 participants with an event per 100 person-years in the abelacimab 90 mg, abelacimab 150 mg, and rivaroxaban groups, respectively. The sponsor’s time-to-event analyses report hazard ratios below 1 for both abelacimab comparisons, with 95% confidence intervals excluding 1 and p < 0.001.
These findings support a statistical interpretation about the defined composite outcome and the comparisons reported. They do not, on their own, establish the magnitude of an individual participant’s absolute risk over a fixed period. They also do not show how the result breaks down between major and CRNM bleeding, because component-specific primary results are not reported in the registry record.
The result does not establish a conclusion about all possible safety outcomes, the secondary endpoints, stroke outcomes, or any unreported outcome. The registry’s serious adverse-event and death counts offer additional descriptive context, but they are not substitutes for the endpoint-specific comparative analysis. Nor do they establish whether individual events were related to treatment.
General statistical interpretation also distinguishes evidence against a null hypothesis from the probability that the result is true or clinically important. The reported p-values do not measure the chance of replication, and the confidence intervals do not remove uncertainty. The meaning of the result remains bounded by the trial population, treatment regimens, analysis population, follow-up, endpoint definition, and statistical methods reported in the registry.
The record does not provide enough detail to evaluate all aspects of analysis conduct, including missing data, censoring, multiplicity handling, or the complete protocol plan. Those omissions limit how far the posted result can be extended beyond the reported comparisons. The appropriate reading is specific: the registry reports lower observed primary composite event rates in the two abelacimab groups than in the rivaroxaban group and reports superiority analyses for each comparison, while important questions about other outcomes and analysis details remain unresolved by the record.
14. Why This Trial Matters Statistically
This trial is statistically notable for presenting both person-time rates and time-to-event comparisons for a first-event bleeding endpoint. The incidence rates communicate the observed event frequency relative to follow-up, while the hazard ratios summarize comparative event timing under the sponsor’s stratified log-rank analyses. Understanding that these statistics answer related but distinct questions helps avoid treating a hazard ratio as a simple rate ratio or an absolute risk difference.
The three-arm design also makes the comparison structure important. The reported inferential analyses compare each abelacimab dose separately with rivaroxaban. They do not provide a direct test of the 90 mg dose against the 150 mg dose. Even when two active comparisons each have reported p-values below 0.001, that does not establish that the two active doses differ from one another.
The composite endpoint brings major and clinically relevant non-major bleeding into one measure. That can make a broader first-event outcome available for comparison, while also meaning that the overall result does not reveal which component contributed most. The separately listed secondary outcomes could address related questions, but their numerical results are not included in the registry record.
The record also illustrates why interpretation depends on trial details beyond a point estimate. The analysis population is randomized participants who received at least one dose, the endpoint uses follow-up time, and the registry does not report several details about missing observations and analysis planning. These features do not invalidate the reported findings, but they define what can be concluded from the registry record and where further information would be needed for a fuller statistical assessment.
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 full protocol or statistical analysis plan. In particular, the rationale for the sample size, the planned analysis population beyond the provided description, the stratification factors, and the handling of missing data and censoring are not reported in the registry record. The population description is truncated, which limits evaluation of how participants who were randomized but not treated were handled beyond the stated difference in counts.
The record reports two primary comparisons and lists two secondary outcomes, but it does not report whether multiplicity was controlled or how the outcomes were ordered for inference. Numerical results for the secondary outcomes are not included in the registry record. The primary incidence rates are posted without confidence intervals, and component-specific primary results are not reported. These omissions restrict assessment of rate precision and the contribution of major versus CRNM bleeding.
The registry describes triple masking, while the official title describes rivaroxaban as open-label. The record does not explain this apparent difference or detail who was masked. The treatment descriptions also include extension regimens, but the primary endpoint concerns the randomized phase. Further information about transitions into the extension and their relevance to the reported analysis is not reported in the registry record.
Safety information is presented as counts and selected serious-event terms. Event attribution, timing, duration, and group-specific frequencies for the listed terms are not reported in the registry record. The posted adverse-event counts should consequently be treated as descriptive rather than as a complete comparative safety analysis.
Finally, the registry status is active, not recruiting, and the primary completion date is listed as 2024-02-15. These record details do not resolve the omissions in the analysis description. Conclusions should remain limited to the posted primary endpoint and the sponsor-reported comparisons.
16. Related Tutorials
17. Related Statistical Calculators
18. Sources
- ClinicalTrials.gov: Safety and Tolerability of Abelacimab (MAA868) vs. Rivaroxaban in Patients With Atrial Fibrillation — NCT04755283.
- Registry record last updated 2025-11-06; status last verified 2025-10.
- Results first posted 2025-07-30.
- 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.
Continue through the Clinical Biostats statistical pathway
Use this trial's endpoint as a starting point for deeper study of effect measures, confidence intervals, and sample-size planning.
19. Record Summary
AZALEA is a randomized, parallel-group Phase 2 trial sponsored by Anthos Therapeutics, Inc. It compares monthly subcutaneous abelacimab 90 mg and 150 mg with oral rivaroxaban in patients with atrial fibrillation. The registry lists enrollment of 1287, triple masking, and active, not recruiting status.
The primary outcome is the incidence rate of a first ISTH-defined major or CRNM bleeding event per 100 person-years during the randomized phase, up to 33 months. The posted rates are 2.64 for abelacimab 90 mg, 3.22 for abelacimab 150 mg, and 8.39 for rivaroxaban. Sponsor-reported stratified log-rank analyses give hazard ratios of 0.314 and 0.382, respectively, with 95% confidence intervals of 0.192 to 0.513 and 0.243 to 0.602; both p-values are < 0.001.
The record lists two secondary bleeding outcomes and descriptive adverse-event counts. It does not report detailed multiplicity procedures, missing-data methods, or a complete statistical analysis plan.