Study to Evaluate the Efficacy and Safety of RZ402 in Diabetic Macular Edema (DME): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Study to Evaluate the Efficacy and Safety of RZ402 in Diabetic Macular Edema (DME) (NCT05712720), a randomized phase 2 study. The primary outcome is Change in Central Subfield Thickness in Study Eye.

2. Clinical Question

The central question is about a measured change in retinal thickness, not directly about how participants saw or felt. Visual acuity and diabetic retinopathy severity were also listed among secondary outcomes, but they address different aspects of the condition. The registry record describes the primary measure as a comparison with placebo and reports estimates for each dose group against that comparator.

The trial enrolled 94 participants overall. The endpoint analysis lists different group counts from the overall enrollment, and those counts should be read as the participants represented in that endpoint result rather than as a replacement for total enrollment. The record does not explain all differences between those counts and enrollment. This distinction matters: the question is what the posted analysis estimates for the Study Eye at the specified assessment, not what happened to every enrolled participant under every possible outcome measure.

3. Trial Design

This was a randomized, double-masked, placebo-controlled, parallel-arm phase 2 study of RZ402 in participants with diabetic macular edema. The listed masking level is quadruple. Participants were assigned to one of three RZ402 dose groups or a placebo group, with the assigned oral tablet taken once daily for 3 months. The registry lists enrollment of 94 participants and the study status as completed.

The primary endpoint was change from baseline in central subfield thickness in the Study Eye at 12 weeks. The record describes measurement by spectral-domain optical coherence tomography. It reports least-squares mean estimates and 95% confidence intervals for the four groups, followed by mixed-model comparisons of each RZ402 group with placebo. The endpoint description notes that only one eye was selected for each participant, using the eye with worse edema when both eyes were eligible; if both had the same level of edema, the right eye was selected.

The registry lists visual acuity and diabetic retinopathy severity among secondary outcomes. These include change in best corrected visual acuity, gains or losses of at least 5 letters, and change in diabetic retinopathy severity score. The posted results also include safety counts. These outcome types give different views of the study: the primary measure is continuous, some secondary measures are counts of participants meeting a threshold, and safety reporting distinguishes serious from other adverse events.

The study started on 2023-02-06 and reached primary completion on 2024-04-11. The registry does not report a sample-size justification, detailed eligibility criteria, or a full statistical analysis plan. The presence of multiple arms and outcomes alone does not establish how the investigators chose the enrollment target or which analyses were prespecified.

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: Of 229 screened participants, 94 met inclusion criteria and were randomized to treatment.

Overall Study

Analysis population for the primary outcome, as defined in the registry: Patients early-terminated and/or missed assessment

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 in Central Subfield Thickness in Study Eye, measured over 12 weeks.

Central subfield thickness is a measurement of retinal thickness in the Study Eye. The registry describes the endpoint as change from baseline, measured in µm using spectral-domain optical coherence tomography, and assessed at 12 weeks. Reporting change rather than only a measurement at follow-up focuses the analysis on how the measurement shifted from each participant’s starting point.

Thickness is an objective imaging measure, which can be useful in a clinical trial because it provides a quantitative outcome that can be compared across randomized groups. Here, the question is whether the estimated change in each RZ402 group differs from the estimated change in the placebo group. A negative net difference means the estimated change was lower in the RZ402 group than in placebo; it does not, by itself, establish the clinical meaning of that difference.

The endpoint is not the same as visual acuity or a participant’s experience. Those outcomes are listed separately as secondary measures. A change in thickness can therefore be interpreted as evidence about the imaging outcome, while the relationship to vision or other aspects of health requires evidence from the corresponding outcomes and suitable clinical context. That context is not fully reported in the registry record.

The analysis reports least-squares means with 95% confidence intervals. These are model-based estimates, rather than simple descriptions of the raw values. The exact model specification and its assumptions are not reported in the registry record, so the endpoint results should be understood through the reported estimates and comparisons without inferring unreported model details.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary endpoint is continuous: it measures change in central subfield thickness in µm. For a continuous outcome, an analysis can summarize each group using an estimated mean change and then compare those estimates across groups. The registry reports least-squares means, a term for model-based group estimates. These estimates are not necessarily the same as a simple arithmetic average of observed participant values.

What does the confidence interval show?

A confidence interval expresses uncertainty around an estimate under the assumptions of the analysis. The 95% confidence intervals listed for the group estimates show a range of values compatible with the observed data and the model. The intervals for the treatment-versus-placebo comparisons show uncertainty about the net difference. A narrower interval would represent greater precision; a wider interval indicates that a broader range of effects remains compatible with the data. The primary intervals are broad, so the point estimates should not be treated as exact quantities.

For the reported net comparisons, zero represents no difference between the estimated change in an RZ402 group and placebo. An interval that excludes zero indicates that the reported estimate is distinguishable from zero at the corresponding interval level, subject to the analysis assumptions. An interval that crosses zero leaves values on both sides of no difference compatible with the data. This is not a statement about whether an effect is clinically important. Clinical interpretation requires context beyond a statistical interval, and such a threshold is not reported in the registry record.

Why use a mixed model?

The registry identifies the primary comparisons as mixed models analyses. In general, mixed models can be used for outcomes measured across groups and may account for patterns of repeated observations or other modeled structure. Their exact form depends on the data and protocol. The registry record does not report the model formula, covariates, covariance assumptions, estimation procedure, or handling of incomplete repeated measurements. Those details should not be inferred from the method label alone.

The reported least-squares means and mean differences are model-based. Consequently, the arm estimates and the net comparisons should be interpreted as the outputs of the posted model, rather than as raw summaries that must match a simple subtraction. The record does not include enough detail to independently reproduce the mixed-model analysis from the posted summary statistics.

How should the secondary binary outcomes be read?

Some secondary outcomes count participants meeting a criterion, such as a gain or loss of at least 5 letters in best corrected visual acuity. For a binary outcome, each participant either meets the criterion or does not. The displayed counts describe how many participants met it in each group. A risk difference expresses the difference between the proportions in two groups; negative values mean a lower proportion in the RZ402 group than in placebo when the comparison is treatment minus placebo. These comparisons are separate from the primary continuous endpoint.

The secondary results include sponsor-reported analyses as well as unadjusted calculations. An unadjusted calculation from summary counts does not necessarily match an analysis that accounts for stratification or other features of the study. The registry identifies Cochran-Mantel-Haenszel analyses for some visual-acuity comparisons. The exact strata and analysis details are not reported in the registry record, so the method label should not be expanded into assumptions about the design.

Why does over-dispersion matter?

In general, over-dispersion occurs when the variability in count data is greater than a model assumes. If it is ignored, uncertainty can be understated and statistical evidence can appear stronger than warranted. This is a general consideration for some count outcomes, not a claim that over-dispersion occurred in this trial. The registry record does not report an over-dispersion assessment or a specific adjustment for it.

The primary outcome is continuous, so the central methodological issues visible are the model-based estimates, their intervals, and the treatment of incomplete data. The binary secondary outcomes have different assumptions and require separate interpretation. Results across outcome types should not be treated as interchangeable.

What do the planning illustrations mean?

The sample-size section is a planning illustration based on an observed effect, not the trial’s original design calculation. It uses a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation; the assumed effect is a stated fraction of the observed effect. These values describe participants per arm for 80% or 90% power under those assumptions. They are not observed power, do not show the probability that the trial’s result is true, and do not establish that the trial was adequately planned. The trial’s own sample-size justification is not reported in the registry record.

10. Analyses Reported by the Sponsor

The registry’s primary comparison results are reported as mixed-model mean differences, with 95% confidence intervals and p-values. Those sponsor-reported analyses are the appropriate reference for the posted primary comparisons. The registry reports estimates for each RZ402 dose against placebo; it does not report a direct comparison among active doses.

The group-level least-squares means and confidence intervals provide useful context, but they are not a substitute for the mixed-model treatment comparisons. A simple recalculation from summary values would be unadjusted and could differ from the sponsor’s model-based analysis. The registry does not flag a discrepancy. Where an unadjusted recalculation is displayed elsewhere on the page, it should therefore be kept distinct from the sponsor’s analysis rather than treated as a correction or replacement.

The record does not include enough model detail to recreate the sponsor analysis independently. This limits assessment of the precise modeling choices, but it does not change what the posted estimates say: the first two dose comparisons have intervals excluding zero, and the interval for the 400 mg comparison includes zero. These are statistical descriptions of the reported endpoint analysis, not conclusions about clinical importance.

11. Multiplicity, Missing Data and Analysis Populations

The trial has several treatment-versus-placebo comparisons for the primary endpoint, as well as multiple secondary outcomes. When many hypotheses are tested, the chance of at least one apparently unusual result can rise. Statistical procedures can address this through prespecified approaches such as adjusting significance thresholds, organizing tests hierarchically, or designating some analyses as exploratory. Which approach was planned for this study is not reported in the registry record.

The primary results include comparisons for three doses against placebo. The registry gives estimates, confidence intervals, and p-values, but does not state whether these p-values or intervals were adjusted for the set of primary comparisons. The secondary outcomes add further comparisons. Without a reported multiplicity plan, nominal evidence from an individual analysis should be interpreted cautiously and should not be treated as if it necessarily represents a study-wide error-controlled conclusion.

Missing-data handling is another important uncertainty. The primary endpoint population is described as “Patients early-terminated and/or missed assessment.” The record does not explain precisely how that population label relates to the analysis, how missing endpoint values were handled, whether all randomized participants were included, or what assumptions were made about why measurements were unavailable. It also does not describe sensitivity analyses addressing alternative missing-data assumptions. These details matter because a model’s estimates can depend on how incomplete observations are treated.

The endpoint arm counts differ from the total enrollment, and counts also vary across the posted outcome measures. Such differences can arise for several reasons, but the reason in this trial is not reported in the registry record. It would be inappropriate to assume that all enrolled participants contributed to every analysis or to infer the effect of unavailable observations.

The registry does not report interim analyses, a detailed statistical analysis plan, or a sample-size justification. The planning illustrations elsewhere on the page are not evidence of the original design rationale. Taken together, the absent details leave uncertainty about the intended control of false-positive findings and the robustness of results to missing data. These limitations do not erase the posted estimates; they define what can and cannot be concluded from the record.

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 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: acute myocardial infarction (4.3%), orthostatic hypertension (4.3%), dizziness (4.2%), hypoglycemia (4.2%).

The safety results report serious adverse events and other adverse events separately for the four groups. Serious adverse events were reported in 0 participants in the 50 mg group, 1 participant in the 200 mg group, 2 participants in the 400 mg group, and 0 participants in the placebo group. The listed counts at risk are 23, 23, 24, and 24, respectively. Other adverse events were reported in 12 participants in each RZ402 group and 11 participants in the placebo group. No deaths were reported in any group.

The registry also lists serious-event terms with percentages, including acute myocardial infarction, orthostatic hypertension, dizziness, and hypoglycemia. These entries should be read as the terms and percentages posted in the record, not as evidence of a causal relationship to treatment. The registry summary does not provide enough detail to assess event timing, severity beyond the serious classification, recurrence, or attribution.

Safety counts from a trial of this size are descriptive. A low count does not rule out uncommon risks, and a difference in counts does not establish that treatment caused an event. The posted figures can describe what was recorded during the study, but they do not by themselves support a comprehensive safety comparison.

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

The primary results describe the estimated change in an imaging measure over 12 weeks and how that change compared between each RZ402 dose group and placebo. The posted model-based comparisons are negative for all three doses. The confidence intervals for the 50 mg and 200 mg comparisons exclude zero, while the interval for the 400 mg comparison includes zero. This is the statistical pattern in the registry record, with uncertainty quantified by the reported intervals.

That pattern does not establish that a particular dose is clinically beneficial, that one dose is preferable to another, or that the result will persist beyond the reported assessment. The record gives no direct active-dose comparison, and the primary endpoint is not a direct measure of vision or quality of life. The listed visual-acuity and retinopathy outcomes are separate measures and should be considered on their own terms.

A confidence interval is not a guarantee that the true value lies within the displayed bounds, and a result that excludes zero does not establish that the effect is important to participants. Conversely, an interval that includes zero does not prove that there is no effect; it indicates that the data and model remain compatible with a range that includes no difference. The interval widths make precision an important part of interpretation.

The trial record is also incomplete for several questions that would shape a fuller assessment: the analysis plan, multiplicity strategy, detailed missing-data methods, sample-size rationale, and model specification are not reported in the registry record. The outcomes should therefore be understood as the posted findings from a completed phase 2 study, rather than as a complete account of all evidence needed to judge clinical relevance or generalizability. Nothing in these results alone provides a basis for individual medical decisions.

14. Why This Trial Matters Statistically

This trial is statistically interesting because it combines a continuous imaging endpoint with several secondary outcomes that measure different features of diabetic macular edema. The primary measure gives a quantitative estimate of change in central subfield thickness, while the visual-acuity outcomes use participant counts and threshold-based definitions. Comparing these outcomes illustrates why statistical evidence for one endpoint does not automatically transfer to another.

The three dose groups also make the distinction between treatment-versus-placebo comparisons and dose-response evidence important. The registry reports each dose against placebo, but does not report a direct comparison among the doses or a dose-trend analysis. Separate estimates can suggest patterns worth investigating, yet they do not establish that increasing dose causes a corresponding change in outcome.

The confidence intervals are central to the reading. The point estimates alone give an incomplete account; the intervals show that the estimated treatment differences retain appreciable uncertainty. This is particularly useful for clinical researchers interpreting a phase 2 study, where estimation and the scale of uncertainty matter alongside whether a comparison crosses a conventional statistical threshold.

Finally, the record makes visible a recurring challenge in registry-based interpretation: the results can be informative while the supporting statistical detail remains limited. The model label and reported estimates are available, but full specifications, missing-data methods, multiplicity handling, and the original enrollment rationale are not. Recognizing these boundaries helps readers distinguish what the results show from what would require additional documentation or later evidence.

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.

Several limitations are specific to interpreting this record. The primary endpoint is reported at 12 weeks, and the registry material does not establish how the results relate to outcomes at later follow-up. The endpoint is an imaging measure; although visual-acuity and severity measures are listed, the findings do not provide a complete numerical account of every secondary outcome. The relationship between a measured thickness change and clinically meaningful changes is not established by the endpoint description alone.

The Study Eye selection rule means the primary analysis concerns one selected eye per participant, not both eyes as separate observations. The eye-selection procedure is described, but full eligibility criteria and the distribution of participants whose Study Eye was selected under each rule are not reported in the registry record. Generalizability to people who would not meet the trial’s eligibility criteria cannot be assessed from the information.

Endpoint group counts are smaller than overall enrollment, and counts differ across outcomes. The registry does not explain all reasons for these differences or provide detailed missing-data analyses. The primary population label refers to participants who early-terminated and/or missed assessment, but does not clarify the exact analysis population or how missing measurements were handled. This leaves uncertainty about the influence of incomplete observations.

The study has four arms and several outcome measures, while the registry does not report the full analysis plan, multiplicity strategy, sample-size justification, or exact mixed-model specification. As a result, readers cannot establish from the record alone which tests were intended to be confirmatory, how family-wise error was handled, or how sensitive the estimates are to modeling choices. These limitations apply to interpretation of the posted record and do not imply that the underlying study lacked procedures that were not reported.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Study to Evaluate the Efficacy and Safety of RZ402 in Diabetic Macular Edema (DME) — NCT05712720.
• Registry record last updated 2026-09-01; status last verified 2026-08.
• Results first posted 2026-09-01.
• 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

The registry describes a completed, randomized, double-masked, placebo-controlled, parallel-arm phase 2 study of oral RZ402 in diabetic macular edema. It lists 94 participants and four groups: 50 mg, 200 mg, and 400 mg RZ402, plus placebo, each taken once daily for 3 months.

The primary endpoint was change from baseline in central subfield thickness in the Study Eye at 12 weeks, measured using spectral-domain optical coherence tomography. The record reports least-squares means and mixed-model comparisons with placebo. It also lists visual acuity, diabetic retinopathy severity, and safety outcomes.

The posted primary comparisons show negative estimated net differences for all three doses, with confidence intervals excluding zero for the 50 mg and 200 mg comparisons and including zero for the 400 mg comparison. Important details, including the complete analysis plan, multiplicity approach, missing-data methods, and sample-size justification, are not reported in the registry record.