A Study to Learn How Well a Higher Amount of Aflibercept Given as an Injection Into the Eye Works and How Safe it is in People With Reduced Vision Due to Swelling in the Macula, Central Part of the Retina Caused by a Blocked Vein in the Retina (Macula Edema Secondary to Retinal Vein Occlusion): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study to Learn How Well a Higher Amount of Aflibercept Given as an Injection Into the Eye Works and How Safe it is in People With Reduced Vision Due to Swelling in the Macula, Central Part of the Retina Caused by a Blocked Vein in the Retina (Macula Edema Secondary to Retinal Vein Occlusion) (NCT05850520), a randomized phase 3 study. The primary outcome is Change From Baseline in BCVA Measured by the ETDRS Letter Score at Week 36.

2. Clinical Question

The study addresses a practical treatment question as well as a visual-acuity question. The registry describes aflibercept 2 mg as standard care administered every 4 weeks, and says that repeated injections can place a burden on patients. The higher-dose regimens were being studied in the context of potentially less frequent treatment, with treatment intervals adjusted according to response. The record does not establish from these data alone how often any individual would receive injections in practice.

The central statistical comparison is not simply whether each group’s average score changed from baseline. It is whether the average change in each 8 mg regimen was sufficiently close to the active-control result under the sponsor’s non-inferiority analysis. The non-inferiority margin and the detailed decision rule are not reported in the registry record, so the margin-based interpretation cannot be independently reconstructed from the posted results alone.

3. Trial Design

NCT05850520 is a randomized, double-masked, active-controlled Phase 3 study with a parallel-group design. The registry lists quadruple masking. Enrollment is recorded as 892, and the study status is completed. The recorded study start is 2023-05-15 and primary completion is 2024-11-07.

The three listed treatment groups are Aflibercept 2q4, Aflibercept 8q8/3, and Aflibercept 8q8/5. The two higher-dose regimens are described as starting with initiation intervals, then extending treatment intervals and making further adjustments according to treatment response. The standard-of-care group received aflibercept 2 mg on standard treatment intervals, with later adjustment based on response. The registry also lists sham and fluorescein among the interventions, but the record details do not specify their procedural roles in enough detail to describe them further.

The primary endpoint was assessed at Week 36. Other recorded outcomes cover visual acuity and active injections at later time points, as well as visual acuity and retinal-fluid measures at Weeks 36 and 64. The full analysis set includes participants randomly assigned to an intervention who were exposed to an active or sham injection at least once. The number analyzed for each result is recorded with the arm-level results.

The registry record does not report the planned allocation ratio, detailed eligibility criteria, the intended original sample-size justification, or the full schedule of visits and assessments. It also does not give the detailed masking procedures or explain how treatment allocation was concealed. Those omissions limit reconstruction of design details beyond the randomized, masked, parallel, active-controlled structure recorded in the registry.

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: In total 1282 participants were screened, of whom 388 participants were screen failures. Overall, 892 participants were randomized and treated. Of these,795 participants completed the treatment phase through Week 64.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Full analysis set (FAS): All participants randomly assigned to study intervention who were exposed to study intervention (active or sham injection) at least once.

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 Change From Baseline in BCVA Measured by the ETDRS Letter Score at Week 36, measured over At Week 36.

The primary endpoint is the change from baseline in best-corrected visual acuity (BCVA), measured by the ETDRS letter score at Week 36. The registry defines the ETDRS letter score as ranging from 0 to 100 letters and states that a higher score means better visual acuity. The reported arm results are least-squares means, with standard errors as the measure of dispersion.

Visual acuity is central to the study question because macular edema can blur vision, and the macula is responsible for fine vision. A standardized letter-score assessment makes visual acuity expressible on a common scale across study groups and allows the comparison to focus on change from baseline rather than on an isolated score at follow-up. In general, measuring change can account for where participants started, although the registry’s detailed modeling and adjustment procedures are not reported in the record.

The main comparison concerns the difference in average change between each higher-dose regimen and the 2 mg active control. A positive or negative difference must be interpreted in light of the stated direction of the score: higher scores indicate better visual acuity. The registry presents these as non-inferiority analyses, which ask whether an intervention is not unacceptably worse than its comparator under a prespecified margin. The margin and its clinical justification are not reported in the registry record. Therefore, the endpoint values and confidence intervals can describe the observed comparison, but they do not by themselves reveal the complete decision rule used by the sponsor.

7. Results Posted in the Registry

• Aflibercept 8q8/3: mean 17.40 versus 17.50 on control; difference -0.10 (95% CI -2.04 to 1.84), standardized effect d = -0.01; z-test p = 0.920.
• Aflibercept 8q8/5: mean 18.30 versus 17.50 on control; difference 0.80 (95% CI -1.01 to 2.61), standardized effect d = 0.07; z-test p = 0.386.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Number of Active Injections From Baseline to Week 64

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard errors.
• Result:
  • Aflibercept 8q8/3: mean 8.50 versus 11.70 on control; difference -3.20 (95% CI -3.48 to -2.92), standardized effect d = -1.86; z-test p = <0.001.
  • Aflibercept 8q8/5: mean 9.50 versus 11.70 on control; difference -2.20 (95% CI -2.48 to -1.92), standardized effect d = -1.27; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Aflibercept 8q8/3 than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Aflibercept 8q8/5 than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Change From Baseline in BCVA Measured by the ETDRS Letter Score at Week 44*

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard errors.
• Result:
  • Aflibercept 8q8/5: mean 18.80 versus 17.80 on control; difference 1.00 (95% CI -0.94 to 2.94), standardized effect d = 0.08; z-test p = 0.312.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Aflibercept 8q8/5 and the control. The interval gives the range of effects that remain compatible with the data.

Change From Baseline in BCVA Measured by the ETDRS Letter Score at Week 64

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard errors.
• Result:
  • Aflibercept 8q8/3: mean 17.80 versus 17.30 on control; difference 0.50 (95% CI -1.58 to 2.58), standardized effect d = 0.04; z-test p = 0.638.
  • Aflibercept 8q8/5: mean 18.10 versus 17.30 on control; difference 0.80 (95% CI -1.42 to 3.02), standardized effect d = 0.06; z-test p = 0.480.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Aflibercept 8q8/3 and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Aflibercept 8q8/5 and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcomes broadly extend the primary question in two directions: whether visual-acuity patterns persist at later follow-up and how the number of active injections differs across regimens. The numerical summaries available for these outcomes show later BCVA changes in a generally similar range across groups, while the active-injection averages are lower in the two 8 mg groups than in the 2 mg group through Week 64. These group-level patterns are consistent with the study’s interest in treatment intervals and vision, but they do not establish what any individual participant experienced.

The registry names additional outcomes, including visual-acuity thresholds, gains in letter score, retinal-fluid findings, and central sub-field thickness. Numerical results for the remaining secondary outcomes are not reported in the registry record details available for this analysis. The summaries therefore cannot support a complete account of every secondary measure or a comprehensive comparison across all outcomes.

Secondary outcomes can help describe whether findings are coherent across related measures, but they do not replace the prespecified primary endpoint. Multiple outcomes and comparisons create more opportunities for apparently notable results to arise by chance. Unless the analysis plan accounts for this multiplicity, individual secondary p-values generally carry less confirmatory weight than the primary analysis. The registry’s approach to multiplicity, including whether any adjustment or hierarchy was used, is not reported in the registry record. Secondary results should consequently be read as supportive or exploratory evidence rather than as independent proof of a treatment effect.

In particular, lower average injection counts do not by themselves establish the clinical implications of a regimen, and similar later visual-acuity averages do not answer every question about durability, safety, or patient experience. The registry does not report enough detail to establish whether secondary analyses were adjusted for baseline factors, missing observations, or multiplicity. Interpretation should remain tied to the reported summaries and their uncertainty.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Number of Active Injections From Baseline to Week 64
• Number of Active Injections From Baseline to Week 36
• Change From Baseline in BCVA Measured by the ETDRS Letter Score at Week 44*
• Change From Baseline in BCVA Measured by the ETDRS Letter Score at Week 64
• Participants Gaining at Least 15 Letters in BCVA From Baseline at Weeks 36 and 64
• Participants Achieving an ETDRS Letter Score of at Least 69 (Approximate 20/40 Snellen Equivalent) at Weeks 36 and 64
• Participants Having no Intraretinal Fluid (IRF) and no Sub-retinal Fluid (SRF) in the Center Subfield at Weeks 36 and 64
• Change From Baseline in Central Sub-field Thickness (CST) at Week 36

9. Statistical Interpretation

10. Statistical Methods Explained

What does a least-squares mean represent?

A least-squares mean is a model-based estimate of a group’s average outcome. In general, it can reflect adjustments specified in a statistical model rather than being only the arithmetic average of the observed values. The Week 36 arm estimates are reported as least-squares means, but the registry record does not state all model terms, covariates, or other implementation details. They should therefore be read as the sponsor’s estimated group-level changes, not as raw averages that can be reconstructed from the posted arm summaries.

The Week 36 arm estimates are 17.5 letters for Aflibercept 2q4, 17.4 letters for Aflibercept 8q8/3, and 18.3 letters for Aflibercept 8q8/5. Their standard errors indicate uncertainty in the estimated means. A standard error describes how precisely a mean is estimated; it is not the spread of individual participants’ scores. The record does not report the individual-level distributions.

How should the group differences be read?

A difference in least-squares means compares the estimated change in one group with the estimated change in the active-control group. The sponsor reports differences of -0.1 letters for 8q8/3 and 0.8 letters for 8q8/5, each versus 2q4. Under the outcome definition, a higher letter score is better. The signs therefore describe the direction of the estimated differences, but should not be used on their own to claim a meaningful advantage. The intervals and the study’s non-inferiority framework are also relevant.

For the primary outcome, the sponsor used a mixed models analysis. In general, mixed models can analyze repeated or otherwise related observations while representing patterns of variation in the data. The registry does not report the model specification, covariance structure, handling of missing visits, or covariates. It is therefore not possible to reproduce the sponsor’s model or determine from the registry record exactly how each design feature was represented.

What is the role of the non-inferiority analysis?

A non-inferiority analysis evaluates whether an intervention’s result is not worse than a comparator by more than a prespecified margin. The margin is an essential part of the interpretation: a confidence interval must be assessed in relation to that threshold, not merely in relation to zero. The registry labels both primary comparisons as non-inferiority analyses, but the margin, its justification, and the detailed testing procedure are not reported in the registry record. The reported label and results can be described, but the full non-inferiority decision cannot be independently verified from the registry entries.

The registry lists a 95% confidence interval and a p-value of < 0.001 for each sponsor analysis. These are the values as recorded, and the p-values belong to the sponsor’s reported non-inferiority analyses. The re-calculated unadjusted p-values are 0.92 and 0.386. Because the registry flags a discrepancy, the sponsor’s analysis takes precedence. The two sets of p-values should not be treated as interchangeable measures of the same analysis.

What does an unadjusted re-calculation do?

The re-calculations compare posted arm-level summaries with a simple, unadjusted method. For the primary outcome, the estimated differences are -0.1 and 0.8 letters. Their intervals are -2.04 to 1.84 and -1.007 to 2.607. The record indicates that the pooled standard deviations used in these calculations are approximate, reconstructed from the reported standard errors. These calculations do not reproduce the sponsor’s mixed model, its adjustments, or its non-inferiority test. They are useful as a transparent summary-statistic comparison, but they are not a substitute for the trial analysis.

In general, a confidence interval presents a range of values compatible with the data and method under repeated sampling assumptions. An interval spanning zero means that a simple difference-from-zero interpretation does not establish a clear direction of difference at that confidence level. It does not settle a non-inferiority question, which depends on a specified margin. Here, both re-calculated primary intervals span zero, and the sponsor’s intervals also span zero. The registry’s non-inferiority conclusions nevertheless must be interpreted using the sponsor’s method and margin, whose details are not reported.

Why are standard errors and standard deviations different?

A standard deviation describes variation among individual observations, while a standard error describes uncertainty in an estimated mean. The primary arm results report standard errors, not standard deviations. The re-calculation process estimates pooled standard deviations approximately from those standard errors, as indicated in the record. This is a simplifying step; it can make the unadjusted interval and planning calculations differ from analyses that use the original participant-level data or the sponsor’s model.

In general, a planning calculation based on a small observed difference relative to the estimated variation can require many participants per arm to detect that difference under a chosen power and significance level. Such calculations are sensitive to the estimated effect and its variability. They are illustrations based on the observed summaries, not evidence about the design choices or analysis methods used by the sponsor.

11. Confidence Intervals and Precision

The sponsor’s Week 36 95% confidence interval for 8q8/3 versus 2q4 runs from -2.0 to 1.9 letters. For 8q8/5 versus 2q4, it runs from -1.1 to 2.7 letters. Each interval includes zero, so the sponsor’s estimates are compatible with differences in either direction under the reported model. Their width also indicates that the estimated differences are not known with letter-level exactness.

The corresponding unadjusted re-calculated intervals are -2.04 to 1.84 letters for 8q8/3 and -1.007 to 2.607 letters for 8q8/5. The small differences between the re-calculated limits and the sponsor’s reported limits reflect that these are different analyses. The record flags a discrepancy, and the sponsor’s analysis takes precedence. The re-calculated intervals should be treated as approximate and should not be substituted for the model-based intervals.

In general, a confidence interval describes statistical uncertainty around an estimated effect under the method used to calculate it. It is not a range containing a fixed proportion of individual participants’ outcomes, and it does not show the probability that a particular treatment effect is true. Precision depends on the amount of information and the variability represented in the analysis. Here, the arm-level standard errors and approximate re-calculations provide summary-level context, but the registry does not supply participant-level data or full model specifications.

For non-inferiority, the key comparison is between the confidence interval and the prespecified margin. Because the margin and its justification are not reported in the registry record, the intervals alone are insufficient to independently assess the non-inferiority criterion. The registry’s reported analysis label and p-values should be understood as sponsor-reported results, not as a decision that can be reconstructed solely by checking whether an interval includes zero.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The sample-size illustrations are expressed as participants per arm, not as total events. The planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is set to the stated fraction of the observed effect. These calculations use the observed effect as their starting point; the trial’s original design assumptions and sample-size justification are not reported in the registry record.

The 80% power column represents the number of participants per arm in a hypothetical study that, under the stated assumptions and effect, would have an 80% chance of detecting that effect using the chosen test procedure. The 90% power column represents the corresponding hypothetical number per arm for a 90% chance. Power is a property of a planned design under assumptions; it is not a measure of whether the completed study’s observed results are reliable.

The rows that plan for 100%, 75%, or 50% of the observed effect show how the illustration changes when the target effect is reduced. Smaller effects are harder to distinguish from random variation, so they require more participants per arm under otherwise unchanged assumptions. This is especially important when the observed difference is small: a planning calculation that uses the full observed effect can look very different from one that targets a smaller fraction.

The figures are not the sponsor’s original sample-size calculation and do not establish what the trial could or could not detect under its actual analysis plan. They are planning illustrations based on posted summary statistics and approximate variability. They are not observed power.

13. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = -0.100; pooled standard deviation σ = 12.065.
2. n = 2 × (1.96 + 0.842)² × 12.065² / -0.100² = 228,491.1, rounded up to 228,492 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
delta <- -0.100000000000001                      # observed difference in means
sigma <- 12.0646792948046                     # pooled standard deviation
n_per_arm <- function(power, frac = 1) {
  ceiling(round(2 * (z_a + qnorm(power))^2 * sigma^2 / (delta * frac)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (t-based, so very slightly larger)
cat("power.t.test:", round(power.t.test(delta = delta, sd = sigma, power = 0.80)$n, 1), "\n")

# Simulation: exact two-sample t-test using simulated means and variances
set.seed(5850520)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 228,492 (80% power), 305,885 (90% power)
 75% of observed effect: n per arm = 406,207 (80% power), 543,796 (90% power)
 50% of observed effect: n per arm = 913,965 (80% power), 1,223,539 (90% power)
power.t.test: 228492 
Simulated power at n = 228,492 per arm: 0.802 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

14. Analyses Reported by the Sponsor

The sponsor’s own primary analyses are mixed models analyses of differences in least-squares means, and the registry labels both comparisons as non-inferiority. The sponsor reports -0.1 letters for Aflibercept 8q8/3 versus Aflibercept 2q4, with a 95% confidence interval from -2.0 to 1.9 and p < 0.001. For Aflibercept 8q8/5 versus Aflibercept 2q4, the reported difference is 0.8 letters, with a 95% confidence interval from -1.1 to 2.7 and p < 0.001.

The simple re-calculations from the posted arm summaries give the same rounded point differences, but produce different interval limits and p-values: 0.92 for 8q8/3 and 0.386 for 8q8/5. These re-calculations are unadjusted, and the pooled variability is approximate. They do not reproduce the sponsor’s mixed model or non-inferiority procedure. The registry explicitly flags a discrepancy. In that circumstance, the sponsor’s own analysis takes precedence for interpreting the reported trial analysis.

The difference in reported and re-calculated p-values should not be resolved by choosing whichever value appears more persuasive. The registry does not provide sufficient details about the non-inferiority margin, testing procedure, or model to reconstruct the sponsor’s result. The sponsor’s reported estimates and analysis classification can be relayed as recorded, while the re-calculations remain a separate descriptive check rather than a competing trial result.

15. Multiplicity, Missing Data and Analysis Populations

The registry identifies the Week 36 BCVA comparison as the primary outcome and describes the sponsor’s primary analyses as non-inferiority analyses. It also lists multiple secondary outcomes and comparisons across treatment groups and follow-up times. Testing many outcomes can increase the chance of finding an apparently notable result by chance if the analysis does not control the overall error rate. In general, a prespecified hierarchy or multiplicity adjustment can address this issue, but the registry record does not report whether either approach was used, or how it was implemented.

The non-inferiority margin, the rationale for its selection, and the detailed decision criteria are also not reported in the registry record. These details are central to interpreting a non-inferiority result. The registry’s classification and sponsor-reported analyses can be stated, but the full inferential basis cannot be reconstructed from the record details.

The primary population is defined as the full analysis set: all randomly assigned participants exposed to study intervention, active or sham, at least once. The record does not state how participants who did not receive an injection, missed a visit, discontinued treatment, or had incomplete outcome measurements were handled in each analysis. It also does not report the methods used to address missing outcome data, assumptions about missingness, or sensitivity analyses. These details matter because different approaches to incomplete follow-up can affect estimates, particularly in non-inferiority studies.

The registry reports arm-level analysis counts, but does not provide a complete account of participant flow, exclusions from each endpoint analysis, or reasons for missing data in the registry record. Interim analyses, stopping rules, and the full statistical analysis plan are not reported in the registry record. The absence of these details is a limit on independent assessment; it is not evidence that a particular procedure was or was not performed.

16. 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 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: endophthalmitis (1.0%), bradycardia (0.7%), diverticulitis (0.7%), pneumonia (0.7%), sepsis (0.7%).

The registry reports serious adverse events for 44 participants in the Aflibercept 2q4 group, 37 in Aflibercept 8q8/3, and 38 in Aflibercept 8q8/5. The respective numbers recorded at risk are 301, 293, and 298. For the combined All Aflibercept 8mg group, the record lists 75 participants with serious adverse events and 591 at risk. The registry also records deaths of 3, 2, and 5 in the three individual treatment groups, and 7 in the combined 8 mg group.

These counts describe what the registry records; they do not show that an event was caused by treatment. The registry entries do not give causal attribution, event timing, severity breakdown beyond the serious-event classification, or exposure-adjusted rates. The record lists Endophthalmitis at 1.0%, and Bradycardia, Diverticulitis, Pneumonia, Sepsis, and Macular hole at 0.7% each among the top serious-event terms. Those terms should be read as recorded event frequencies, not as comparative evidence of risk without further context.

Other adverse events are also counted in the registry, but a count alone does not describe event type or clinical importance. The serious-event counts should be interpreted alongside the study design, follow-up, and completeness of reporting; those details are not fully reported in the registry record.

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

The primary endpoint shows substantial average improvement from baseline in all three groups, with Week 36 least-squares mean changes of 17.5 letters for Aflibercept 2q4, 17.4 letters for Aflibercept 8q8/3, and 18.3 letters for Aflibercept 8q8/5. The sponsor’s estimated differences between each 8 mg regimen and the active control are small in magnitude relative to the observed group changes. The confidence intervals reflect uncertainty around those estimated differences and extend to values on both sides of zero.

The registry reports the primary comparisons as non-inferiority analyses. That classification is not the same as a demonstration that the regimens are identical, and it is not a test of superiority. A non-inferiority conclusion depends on a prespecified margin and analysis procedure. Because the margin and its rationale are not reported in the registry record, readers cannot independently determine from the intervals alone whether the criterion was met. The sponsor’s analysis takes precedence over the unadjusted re-calculations, which are not a reproduction of the sponsor’s model.

The posted secondary summaries add information about later visual-acuity measures and injection counts. They do not establish how a particular person’s vision or treatment schedule would change. Group means can conceal variation among participants, and the registry entries do not provide participant-level results or a complete analysis of all listed secondary measures. Serious-event counts likewise describe recorded events, not causation or an individual’s risk.

These results should be understood as statistical findings from a randomized trial record, not as medical advice or a recommendation for treatment. The registry record does not contain all the design and analysis details needed to judge every aspect of the evidence. In particular, the non-inferiority margin, multiplicity plan, missing-data methods, and detailed model specifications are not reported. The conclusions that can be stated most securely are the posted group estimates and sponsor-reported analysis labels, along with the uncertainty and reporting limitations surrounding them.

18. Why This Trial Matters Statistically

This trial is statistically informative because it compares two higher-dose aflibercept regimens with an active standard-of-care group on a defined visual-acuity endpoint. A randomized, masked, parallel-group design is intended to support a structured comparison while reducing some sources of bias. The use of an active control is relevant because the study question concerns the performance of a higher-dose regimen relative to an established treatment approach described in the registry.

The primary result also illustrates why the statistical question matters as much as the observed group means. Similar estimates do not by themselves prove equivalence or non-inferiority. Non-inferiority depends on an agreed margin, appropriate analysis, and adequate precision. The registry labels the comparisons as non-inferiority, but does not report the margin or full analysis plan. This makes the distinction between sponsor-reported conclusions and independent reconstruction especially important.

The study also illustrates the difference between model-based trial results and summary-statistic re-calculations. The re-calculations offer a transparent unadjusted comparison, but they do not reproduce the sponsor’s mixed model. With a recorded discrepancy, the sponsor’s own analysis takes precedence, while the mismatch remains a reason to interpret the re-calculated values cautiously.

Finally, the injection-count outcomes provide a statistical complement to visual-acuity outcomes: they describe treatment use as well as measured vision. Taken together, the outcomes show why a trial’s interpretation requires attention to endpoint hierarchy, analysis populations, precision, multiplicity, and the completeness of registry reporting—not just the direction of an estimated difference.

19. 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 does not report the non-inferiority margin or the justification for it. Without that information, readers cannot independently evaluate whether the sponsor’s confidence intervals satisfy the prespecified criterion. The sponsor’s analyses are identified as non-inferiority analyses, but the full statistical basis is not available in the record details.

There is a flagged discrepancy between the sponsor’s reported primary analyses and simple unadjusted re-calculations. The re-calculations use approximate pooled standard deviations inferred from standard errors and do not reproduce the mixed models analysis. The sponsor’s own analysis takes precedence, but the discrepancy limits the value of the re-calculations as a check on the primary inference.

The detailed statistical model, including its covariates and other assumptions, is not reported in the registry record. Nor are the missing-data methods, endpoint-specific analysis populations, multiplicity strategy, or interim-analysis procedures. These omissions limit independent assessment of robustness and of how the analyses account for repeated measurements and incomplete follow-up.

The available secondary numerical results cover only a portion of the listed outcomes. The registry entries do not provide numerical summaries for the remaining secondary measures, so the overall pattern across all planned outcomes cannot be assessed. The study also reports group-level results; participant-level distributions and the range of individual outcomes are not available in the record details.

Safety counts lack causal attribution and detailed event timing in the registry record for this analysis. The adverse-event summaries therefore cannot independently establish treatment-related risk differences. These reporting limits do not negate the posted results, but they set boundaries on the conclusions that can be drawn from the registry record.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: A Study to Learn How Well a Higher Amount of Aflibercept Given as an Injection Into the Eye Works and How Safe it is in People With Reduced Vision Due to Swelling in the Macula, Central Part of the Retina Caused by a Blocked Vein in the Retina (Macula Edema Secondary to Retinal Vein Occlusion) — NCT05850520.
• Registry record last updated 2026-06-17; status last verified 2026-06.
• Results first posted 2025-12-18.
• 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.

23. Record Summary

NCT05850520 is a completed randomized, double-masked, active-controlled, parallel-group Phase 3 study of aflibercept 8 mg regimens versus aflibercept 2 mg for macular edema secondary to retinal vein occlusion. Enrollment is recorded as 892. The primary endpoint is change from baseline in BCVA measured by ETDRS letter score at Week 36, with higher scores indicating better visual acuity.

The sponsor reports mixed-model non-inferiority analyses for both higher-dose regimens. The registry also contains arm-level estimates and summary information that permit unadjusted re-calculations, but it flags a discrepancy; the sponsor’s own analysis takes precedence. The record does not report the non-inferiority margin, full analysis plan, multiplicity approach, or missing-data methods. Secondary and safety findings should therefore be interpreted as registry-reported group summaries with these limitations in view.