A Study to Evaluate the Efficacy and Safety of Faricimab in Participants With Macular Edema Secondary to Branch Retinal Vein Occlusion: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study to Evaluate the Efficacy and Safety of Faricimab in Participants With Macular Edema Secondary to Branch Retinal Vein Occlusion (NCT04740905), a randomized phase 3 study. The primary outcome is Part 1: Change From Baseline in Best Corrected Visual Acuity (BCVA) in the Study Eye at Week 24.

2. Clinical Question

The trial record describes a Phase III, multicenter, randomized, active comparator-controlled study. The primary comparison concerns Part 1, when the randomized groups received different active treatments. The question is about a measured change in visual acuity, not whether participants experienced every possible effect of treatment or whether the groups were identical in all respects.

The study also followed participants after Week 24. During Part 2, both randomized groups received faricimab according to a personalized treatment interval regimen. That phase addresses a different treatment context from the initial randomized comparison: after the switch, the original aflibercept assignment no longer represents continued aflibercept treatment. The primary question therefore remains the Week 24 contrast between the randomized Part 1 groups.

3. Trial Design

This was a Phase III, randomized, parallel-group study of faricimab and aflibercept in participants with macular edema secondary to branch retinal vein occlusion. The registry record lists 553 participants enrolled and a completed study. The study began on 2021-03-02, and primary completion was recorded as 2022-07-06. Hoffmann-La Roche is listed as the sponsor.

In Part 1, participants assigned to Arm A received faricimab 6 mg by intravitreal injection every 4 weeks from Day 1 through Week 20. Participants assigned to Arm B received aflibercept 2 mg by intravitreal injection on the same schedule. The primary endpoint was assessed at Week 24. The primary analysis included 276 participants in Arm A and 277 in Arm B.

Part 2 ran from Week 24 to Week 72. Participants in both original groups received faricimab under a personalized treatment interval regimen. Consequently, the randomized active-comparator contrast belongs to Part 1; Part 2 does not preserve a continuing faricimab-versus-aflibercept comparison.

The official title describes the study as double-masked, while the registry masking classification is triple. The record also describes an active comparator-controlled study and lists a sham procedure among the interventions. Further details explaining the masking classification and the role of the sham procedure are not reported in the registry record. The record describes randomized assignment and parallel groups, but does not report the allocation ratio or additional operational details about concealment.

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.

Part 1 (Baseline up to Week 24)

Reasons for not completing this period:

Part 2 (Week 24 to Week 72)

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Intent-to-Treat (ITT) Population: all participants who were randomized in the study, grouped according to the treatment assigned at randomization.

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: ITT Population: All global participants who were randomized in the study, according to the assigned treatment.

6. Primary Endpoint

The registry lists the primary outcome as Part 1: Change From Baseline in Best Corrected Visual Acuity (BCVA) in the Study Eye at Week 24, measured over From Baseline through Week 24.

The primary endpoint was change from baseline in best corrected visual acuity in the study eye at Week 24, measured in ETDRS letters. The ETDRS chart was administered at a starting distance of 4 meters. The score ranges from 0 to 100, with a higher score indicating better visual acuity; a gain from baseline indicates improvement in the measured score.

A change score is useful because it relates the Week 24 measurement to each participant’s baseline rather than describing only the final score. The trial reports mean change for each randomized group and an adjusted difference between those means. The analysis adjusted for treatment arm, visit, their interaction, baseline visual acuity, and randomization stratification factors. This makes the primary estimate a model-based comparison rather than a simple subtraction of the two posted group means.

Visual acuity is a continuous measure, so the mean captures average change across the analyzed participants. It does not describe the full spread of individual responses or establish that every participant changed by the group average. The confidence interval communicates uncertainty around the estimated difference. The registry record reports the primary endpoint for the Intent-to-Treat population, grouped according to randomized assignment.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary endpoint is continuous: each participant has a change in ETDRS letter score, and the analysis compares mean changes between randomized groups. The registry reports an adjusted mean difference with a 95% confidence interval. A confidence interval describes the precision of an estimate under the analysis assumptions; it is not a range containing every participant’s response, nor does it mean that the true value has a stated probability of falling inside this particular interval.

Why use a repeated-measures model?

The sponsor used a Mixed Model of Repeated Measures (MMRM). In general, this type of model analyzes measurements collected at multiple visits while accounting for their relationship within a participant. It can estimate treatment differences over time while adjusting for specified baseline and design factors. The registry describes categorical terms for treatment arm, visit, and the visit-by-treatment-arm interaction, along with baseline BCVA as a continuous covariate. It also describes randomization stratification factors as fixed effects.

The specified stratification factors were baseline BCVA categories of at least 55 and no more than 54 letters, and region categories covering the U.S. and Canada, Asia, and the rest of the world. The registry record reports an unstructured covariance structure. In general, a covariance structure represents how repeated measurements are related; an unstructured choice allows the model to estimate those relationships without imposing a simpler pattern. Whether that choice was appropriate in practice cannot be judged from the reported estimate alone.

What does adjustment change?

An adjusted mean difference is a model-based contrast that accounts for the covariates named in the analysis description. It is distinct from comparing the two group means without adjustment. This distinction matters when interpreting the posted arm summaries: their difference need not reproduce the adjusted treatment contrast. The estimate of -0.6 letters and its 95% confidence interval from -2.2 to 1.1 are the sponsor’s adjusted comparison, not an unadjusted calculation from the arm means.

The primary analysis population was the Intent-to-Treat population: all randomized participants, analyzed according to the treatment assigned at randomization. This preserves the randomized group comparison even if treatment was discontinued or another intercurrent event occurred. The registry says a treatment-policy strategy was applied to all observed values for discontinuation due to adverse events or lack of efficacy and for use of prohibited therapy. In general, a treatment-policy strategy aims to interpret the treatment effect using outcomes regardless of those events, rather than treating the events as reasons to redefine the randomized groups.

How were missing measurements handled?

The registry states that missing data were implicitly imputed by the MMRM under a missing-at-random assumption. In general, this assumption means that, conditional on information included in the model, the probability that an outcome is missing does not depend on the unobserved value itself. It is an assumption, not a fact demonstrated by the model. If missingness depends on unobserved outcomes even after accounting for modeled information, the resulting estimate may be affected. The numbers and timing of missing measurements, and sensitivity analyses addressing departures from this assumption, are not reported in the registry record.

The 95% confidence interval is described as a rounding of a 95.03% interval. The reported interval uses the rounded 95% label. Rounding affects the displayed precision, not the underlying clinical outcome. The registry reports a superiority p-value of 0.498 and separately labels a non-inferiority analysis; the non-inferiority margin and its decision rule are not reported in the registry record.

Why do the analyses answer different questions?

The superiority analysis tests whether the data provide evidence of a difference under its specified null hypothesis. A non-inferiority analysis instead evaluates whether a treatment is not worse than a comparator by more than a prespecified margin. The margin is essential to interpret that assessment and is not reported in the registry record. A confidence interval that spans no difference does not, without the margin and the planned decision rule, resolve the non-inferiority question.

Multiplicity is relevant because the registry lists 48 outcome measures and 4 analyses, including a primary endpoint and numerous secondary outcomes. Testing multiple outcomes or contrasts can increase the chance of at least one apparently unusual result when all null hypotheses hold. General methods for addressing this include prespecified testing sequences or other error-control procedures. The procedure used for the secondary outcomes is not reported in the registry record, so individual secondary findings should not automatically be treated as confirmatory.

10. Analyses Reported by the Sponsor

The sponsor-reported primary comparison is a Mixed Model of Repeated Measures estimate: the difference in adjusted means was -0.6 ETDRS letters, with a 95% confidence interval from -2.2 to 1.1. The superiority analysis reports p = 0.498. The separately labeled non-inferiority analysis reports the same point estimate and confidence interval, but no p-value is listed for that analysis.

No unadjusted recalculation of the primary continuous contrast is available from the posted summary statistics. The arm-level mean changes are 16.9 letters for faricimab and 17.5 letters for aflibercept, but subtracting those displayed means would not reproduce the adjusted MMRM contrast. The sponsor’s estimate accounts for the analysis terms described in the registry record; it should therefore be read as the principal reported comparison rather than as a raw difference between summaries.

The record does not flag a discrepancy between the available result and a recalculation. The remaining issue is interpretive rather than numerical: the non-inferiority margin and decision rule are not reported in the registry record. The reported confidence interval and superiority p-value describe the estimate and superiority test, but do not fill in those missing non-inferiority details.

11. Multiplicity, Missing Data and Analysis Populations

The trial has a specified primary endpoint and a broad set of secondary outcomes. The registry record lists 48 outcome measures and 4 analyses, while the secondary outcomes include multiple visual-acuity measures at different timepoints and multiple thresholds for gains or avoided losses. When many outcomes are tested, the chance of at least one apparently significant result can increase. The statistical plan for controlling that risk, including whether testing was hierarchical or adjusted in another way, is not reported in the registry record.

The primary result is an adjusted MMRM estimate for the Intent-to-Treat population. The registry describes all randomized participants as grouped by assigned treatment. It also states that all observed values were used under a treatment-policy strategy for discontinuation due to adverse events or lack of efficacy and for use of prohibited therapy. These details clarify how those events were handled conceptually; they do not show how many participants experienced them or how many measurements were missing.

Missing data were implicitly imputed by the MMRM under a missing-at-random assumption. In general, this assumption can be difficult to verify from observed data alone. The count of missing Week 24 values, reasons for missingness, and sensitivity analyses under alternative missing-data assumptions are not reported in the registry record. These omissions limit assessment of how dependent the result may be on the model’s missing-data assumption.

The registry labels a non-inferiority analysis, but the margin and the decision rule are not reported. The record also does not report a sample-size justification, interim analyses, or the full statistical analysis plan. Those details would help establish how the primary and secondary analyses were prespecified and how the planned precision related to the study design. Their absence does not change the reported estimate, but it restricts evaluation of the broader testing strategy.

12. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 3% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: cerebral infarction (1.1%), retinal ischaemia (0.7%), cerebrovascular accident (0.7%), ischaemic stroke (0.7%), myocardial infarction (0.7%).

The registry reports serious adverse events and other adverse events for the two treatment periods. In Part 1, serious adverse events affected 12 of 276 participants in the faricimab group and 17 of 274 in the aflibercept group. Other adverse events affected 62 of 276 and 69 of 274, respectively. Deaths were recorded for 1 participant in the faricimab group and 0 in the aflibercept group during Part 1.

In Part 2, after participants in both original groups received faricimab under the personalized treatment interval regimen, serious adverse events affected 29 of 270 participants in the group originally assigned to faricimab and 26 of 267 in the group originally assigned to aflibercept. Other adverse events affected 92 of 270 and 87 of 267, respectively. Deaths were recorded for 1 and 2 participants in those groups.

These are counts of affected participants, not measures of causal attribution or exposure-adjusted risk. Part 2 is a different treatment context from the randomized active-comparator period, so its groups should not be interpreted as continuing faricimab-versus-aflibercept arms. The registry lists cerebral infarction at 1.1%, and retinal ischaemia, cerebrovascular accident, ischaemic stroke, myocardial infarction, and unstable angina at 0.7% each among the top serious-event terms. The record does not establish from these counts whether an event was caused by treatment.

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

The primary result estimates the adjusted difference between randomized groups in mean change in best corrected visual acuity at Week 24. The point estimate is -0.6 letters, and the 95% confidence interval extends from -2.2 to 1.1. This conveys both the estimated direction and the uncertainty around it. It does not establish that the treatments have identical effects, because a range of differences remains compatible with the reported interval.

The result also does not establish that one treatment is superior. The reported superiority p-value is 0.498, and the confidence interval spans values on either side of no difference. Nor does that result alone establish non-inferiority. The registry labels a non-inferiority analysis, but its margin and decision rule are not reported in the registry record. A non-inferiority conclusion depends on those prespecified details.

The estimate applies to the specified visual-acuity outcome, population, analysis, and Week 24 timepoint. It is a group-level mean comparison, not a prediction of an individual participant’s change. It does not answer every question about safety, other outcomes, or later follow-up. In particular, participants in both groups received faricimab during Part 2, so the Week 24 comparison should not be extended to claim a continued randomized comparison of faricimab and aflibercept through Week 72.

Finally, a lack of clear evidence of superiority is not proof that there is no meaningful difference. The confidence interval describes uncertainty, and the registry record does not report a threshold for interpreting clinical importance. The results are best read as a comparatively precise estimate of the primary endpoint under the stated model, with limits on the conclusions that can be drawn from the missing non-inferiority details and the incomplete secondary-outcome results.

14. Why This Trial Matters Statistically

This trial illustrates why the statistical estimand and the design timeline matter. The randomized comparison is between faricimab and aflibercept during Part 1, with a Week 24 primary endpoint. After that time, both original groups received faricimab under a personalized treatment interval regimen. Treating the later period as if it retained the original active comparison would misstate what randomization can establish.

It also shows the distinction between a continuous primary outcome and threshold-based secondary outcomes. Mean change uses the full scale of observed change, whereas a threshold outcome classifies participants according to whether they cross a specified cutoff. These measures can provide different perspectives, but secondary analyses require appropriate caution when many outcomes are examined and the multiplicity plan is not reported.

The adjusted model and the group means answer related but distinct questions. The MMRM accounts for repeated measurements and prespecified covariates, while the displayed arm means are summaries. The adjusted contrast and confidence interval should therefore be interpreted as the model-based result. The estimate is close to no difference, but the confidence interval communicates uncertainty that a single point estimate cannot.

Finally, the registry’s non-inferiority label makes the missing margin especially important. A non-inferiority claim cannot be assessed merely by observing that the superiority test is not significant. The margin and planned decision rule define the question. This is a useful example of why clear reporting of the analysis objective, missing-data assumptions, multiplicity plan, and sample-size rationale is essential to understanding what a trial’s results support.

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 primary result concerns visual acuity at Week 24. Although the study continued through Week 72, both original groups received faricimab in Part 2, so the later phase does not provide a continuing randomized comparison between faricimab and aflibercept. The available information does not include later endpoint estimates, limiting interpretation of longer-term outcomes.

The registry reports the MMRM approach and its missing-at-random assumption, but does not report the number of missing primary endpoint measurements, reasons for missingness, or sensitivity analyses. The extent to which the estimate depends on the missing-data assumption therefore cannot be assessed from the registry record. The non-inferiority margin and decision rule are also not reported, preventing a full assessment of the non-inferiority analysis.

The record lists numerous secondary outcomes, but endpoint-specific results are not available for most of them in the information described. The multiplicity strategy is not reported, so the confirmatory status of secondary findings cannot be determined. A sample-size justification and interim analyses are not reported in the registry record either.

Safety counts are presented for the treatment periods, but the information does not provide exposure-adjusted comparisons or establish causal attribution. The Part 2 groups reflect different original assignments but receive the same treatment regimen during that period. In addition, the registry classification lists triple masking while the official title says double-masked; details explaining that difference are not reported in the registry record. These reporting limits should be kept separate from the primary estimate itself.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: A Study to Evaluate the Efficacy and Safety of Faricimab in Participants With Macular Edema Secondary to Branch Retinal Vein Occlusion — NCT04740905.
• Registry record last updated 2024-08-06; status last verified 2024-07.
• Results first posted 2024-01-18.
• 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

NCT04740905 is a completed Phase III randomized, parallel-group study sponsored by Hoffmann-La Roche. It enrolled 553 participants with macular edema secondary to branch retinal vein occlusion. Part 1 compared faricimab with aflibercept, each administered every 4 weeks, and the primary endpoint was change from baseline in best corrected visual acuity at Week 24.

The primary MMRM estimate was a difference in adjusted means of -0.6 ETDRS letters, with a 95% confidence interval from -2.2 to 1.1 and a reported superiority p-value of 0.498. Part 2 followed both original groups on faricimab under a personalized treatment interval regimen. The registry also reports adverse-event counts and lists secondary outcomes, but the multiplicity plan, non-inferiority margin, sample-size justification, and several endpoint-specific results are not reported in the registry record.