A Study to Test Efficacy and Safety of Rozanolixizumab in Adult Patients With Generalized Myasthenia Gravis: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study to Test Efficacy and Safety of Rozanolixizumab in Adult Patients With Generalized Myasthenia Gravis (NCT03971422), a randomized phase 3 study. The primary outcome is Change From Baseline to Day 43 in Myasthenia Gravis-Activities of Daily Living (MG-ADL) Score.

2. Clinical Question

The registry record identifies the study as a phase 3, randomized, placebo-controlled trial in generalized myasthenia gravis. The primary endpoint concerns a patient-reported measure of daily functioning and symptoms. The posted primary estimates are least-squares mean changes for each group and comparisons of each rozanolixizumab group with placebo.

This question is about the measured difference between groups at the specified endpoint and the uncertainty around that estimate. The primary outcome is not a direct measure of every aspect of the condition, nor does it by itself describe durability beyond Day 43. Secondary outcomes address additional measures, including response on MG-ADL and changes in other myasthenia gravis scales. Their results provide complementary evidence, while the primary endpoint remains the principal outcome named in the record.

3. Trial Design

The registry record describes a completed phase 3, randomized, parallel-group study sponsored by UCB Biopharma SRL. The listed enrollment is 200. The record gives a study start date of 2019-06-03 and a primary completion date of 2021-08-31. The official title describes a double-blind, placebo-controlled study; the masking field lists quadruple masking.

Participants were assigned to placebo or one of two rozanolixizumab dosage groups, labelled approximately 7 mg/kg and approximately 10 mg/kg in the posted outcome results. Arm descriptions state that assigned rozanolixizumab doses were given at prespecified times during the treatment period, but the timing and detailed regimen are not reported in the registry record. The placebo group received placebo.

The primary endpoint was change from baseline to Day 43 in MG-ADL score. The primary analysis population is the Randomized Set, defined in the record as participants who were randomized and analyzed according to assigned treatment rather than treatment received. The primary outcome estimates include 67 participants in placebo, 66 in the approximately 7 mg/kg group, and 67 in the approximately 10 mg/kg group. Safety summaries use different group counts, which are described with the safety results.

The record lists secondary outcomes covering MG-ADL response, changes in MG-C and QMG scores, patient-reported symptom measures, and treatment-emergent adverse events. It lists 9 outcome measures and 14 analyses. Details of the original sample-size justification, randomization procedure, treatment-period duration, interim analyses, and full statistical analysis plan are not reported in the registry record.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Before assignment: Participant Flow refers to the Randomized Set. Two participants randomized to RLZ \~7mg/kg, were administered RLZ \~10mg/kg at baseline visit. So, these two participants were included in RLZ \~7 mg/kg group in randomized set, but in RLZ \~10 mg/kg group in safety set.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The Randomized Set consisted of all study participants who were randomized and analyzed according to the treatment assigned instead of the actual treatment received.

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: Baseline Characteristics refers to the Randomized Set (RS) which consisted of all study participants who were randomized and analyzed according to the treatment assigned instead of the actual treatment received.

6. Primary Endpoint

The registry lists the primary outcome as Change From Baseline to Day 43 in Myasthenia Gravis-Activities of Daily Living (MG-ADL) Score, measured over Baseline and Day 43.

The primary endpoint is the change from baseline to Day 43 in the Myasthenia Gravis Activities of Daily Living score, abbreviated MG-ADL. The record describes MG-ADL as an 8-item patient-reported instrument developed on the basis of the Quantitative Myasthenia Gravis measure. Its items cover ocular, bulbar, respiratory, and axial symptoms and disability. Each item is graded from 0 to 3, and the total score ranges from 0 to 24.

Higher MG-ADL scores indicate more disability. The direction of change is therefore important: a negative change indicates improvement, while a positive change indicates worsening. The posted group estimates are least-squares mean changes, accompanied by standard errors. The comparisons assess the difference in mean change between each rozanolixizumab group and placebo.

A patient-reported daily-living measure is relevant because it captures symptoms and functional effects as reported by participants, rather than relying only on an examination-based score. The endpoint offers a defined way to compare groups over the stated baseline-to-Day-43 period. It does not capture every aspect of myasthenia gravis, and the Day 43 measurement alone does not establish how scores may change at other times. The registry record does not report a threshold for what constitutes a clinically meaningful change.

7. Results Posted in the Registry

• Rozanolixizumab ~7 mg/kg: mean -3.37 versus -0.78 on control; difference -2.59 (95% CI -3.94 to -1.24), standardized effect d = -0.65; z-test p = <0.001.
• Rozanolixizumab ~10 mg/kg: mean -3.40 versus -0.78 on control; difference -2.62 (95% CI -3.98 to -1.26), standardized effect d = -0.65; z-test p = <0.001.

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.

Percentage of Participants Achieving Myasthenia Gravis-Activities of Daily Living (MG-ADL) Response at Day 43

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Rozanolixizumab ~7 mg/kg: 68.2% (45/66) versus 28.4% (19/67) on control; risk difference 39.8 percentage points (95% CI 24.2 to 55.4), risk ratio 2.40 (1.59–3.64); unadjusted two-proportion z-test p = <0.001.
  • Rozanolixizumab ~10 mg/kg: 61.2% (41/67) versus 28.4% (19/67) on control; risk difference 32.8 percentage points (95% CI 16.9 to 48.7), risk ratio 2.16 (1.41–3.30); unadjusted two-proportion z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for Rozanolixizumab ~7 mg/kg 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 proportion with the event is statistically higher for Rozanolixizumab ~10 mg/kg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Change From Baseline to Day 43 in Myasthenia Gravis-Composite (MG-C) Total Score

• 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:
  • Rozanolixizumab ~7 mg/kg: mean -5.93 versus -2.03 on control; difference -3.90 (95% CI -6.44 to -1.36), standardized effect d = -0.52; z-test p = 0.003.
  • Rozanolixizumab ~10 mg/kg: mean -7.55 versus -2.03 on control; difference -5.53 (95% CI -8.09 to -2.96), standardized effect d = -0.73; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Rozanolixizumab ~7 mg/kg 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 Rozanolixizumab ~10 mg/kg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Change From Baseline to Day 43 in Quantitative Myasthenia Gravis (QMG) Total Score

• 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:
  • Rozanolixizumab ~7 mg/kg: mean -5.40 versus -1.92 on control; difference -3.48 (95% CI -5.37 to -1.60), standardized effect d = -0.63; z-test p = <0.001.
  • Rozanolixizumab ~10 mg/kg: mean -6.67 versus -1.92 on control; difference -4.76 (95% CI -6.66 to -2.85), standardized effect d = -0.85; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Rozanolixizumab ~7 mg/kg 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 Rozanolixizumab ~10 mg/kg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The secondary outcomes address several aspects of generalized myasthenia gravis beyond the primary MG-ADL change score. They include MG-ADL response at Day 43, changes in MG-C and QMG total scores, patient-reported symptom domains, and treatment-emergent adverse events. The numerical outcome summaries available for interpretation include MG-ADL response and changes in MG-C and QMG. Together, these measures provide related but distinct perspectives; they should not be treated as interchangeable versions of the primary endpoint.

The overall pattern in the numerical secondary results is directionally consistent with the primary result: the posted comparisons favor larger changes or greater response in the rozanolixizumab groups than in placebo, and the sponsor-reported comparisons have p-values below 0.05. That pattern offers supporting context for the primary MG-ADL finding, but it does not independently establish the same conclusion for every listed outcome. The registry lists additional patient-reported symptom and safety outcomes, and their detailed numerical estimates are not reported in the registry record.

Secondary outcomes commonly serve as supportive or exploratory evidence, particularly when a trial evaluates multiple measures and comparisons. If many hypotheses are tested, the chance of obtaining at least one small p-value can increase. The registry record does not report a multiplicity adjustment or testing hierarchy, so unadjusted secondary p-values carry less weight than a prespecified primary analysis. They should be considered with their estimates, uncertainty, clinical measurement context, and the number of outcomes assessed, rather than read as a collection of independent confirmations.

The relationship between the primary and secondary outcomes is also limited by timing and measurement. Several outcomes are assessed at Day 43, but they measure different constructs or use different score definitions. Consistency across related measures can be informative, while differences between them may reflect what each instrument captures. The registry record does not supply enough information to infer how strongly the outcomes were expected to correlate or how the investigators planned to interpret discordant results.

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

• Percentage of Participants Achieving Myasthenia Gravis-Activities of Daily Living (MG-ADL) Response at Day 43
• Change From Baseline to Day 43 in Myasthenia Gravis-Composite (MG-C) Total Score
• Change From Baseline to Day 43 in Quantitative Myasthenia Gravis (QMG) Total Score
• Change From Baseline to Day 43 in the Myasthenia Gravis (MG) Symptoms Patient Reported Outcome (PRO) 'Muscle Weakness Fatigability' Score
• Change From Baseline to Day 43 in the Myasthenia Gravis (MG) Symptoms Patient Reported Outcome (PRO) 'Physical Fatigue' Score
• Change From Baseline to Day 43 in the Myasthenia Gravis (MG) Symptoms Patient Reported Outcome (PRO) 'Bulbar Symptoms' Score
• Number of Participants With Treatment-emergent Adverse Events (TEAEs)
• Number of Participants With Treatment-emergent Adverse Events (TEAEs) Leading to Withdrawal of Investigational Medicinal Product (IMP)

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is continuous: each participant has an MG-ADL score, and the posted summaries describe mean change from baseline to Day 43. A mean summarizes the central level of a group’s measurements. The trial reports least-squares means, which are model-based group means; details of the full model specification are not reported in the registry record. The reported standard errors describe the precision of those group mean estimates, rather than the spread of individual participant scores.

The reported primary contrasts are least-squares mean differences comparing each rozanolixizumab dosage group with placebo. The sponsor’s record names the Lehmacher and Wassmer method for these analyses and reports a 95% confidence interval and p-value for each contrast. A confidence interval gives a range of estimates compatible with the data under the method’s assumptions. It is not a range containing a specified proportion of individual participant outcomes, and it does not give the probability that the true effect lies in the interval.

The recalculated contrasts are unadjusted comparisons based on the posted group means and standard errors. The associated pooled standard deviations are approximate, as indicated in the calculation details. These recalculations are useful as a transparent check on the scale and direction of the posted differences, but they do not reproduce all features of a fitted trial analysis. They do not incorporate any model adjustment or other analysis choices that are not represented in the group summaries. The sponsor-reported analysis is the primary account of the trial’s own results.

For the binary secondary outcome, MG-ADL response at Day 43, the posted summaries give the percentage of participants meeting the recorded response outcome in each group. A risk difference expresses the difference between two group proportions. The recalculated risk differences are 0.398 for approximately 7 mg/kg versus placebo and 0.328 for approximately 10 mg/kg versus placebo. The sponsor’s reported analysis uses a Wald test for an odds ratio and reports 95% confidence intervals and p-values. An odds ratio compares the odds of the outcome between groups; it is not the same quantity as a difference in percentages. The sponsor-reported odds ratios are 5.765 and 4.273, respectively, each with p < 0.001.

For continuous secondary outcomes, the record reports least-squares mean changes and group comparisons. The recalculated mean differences use posted group summaries and are unadjusted; the registry’s own comparison method is again listed as Lehmacher and Wassmer. For MG-C, the sponsor reports differences of -3.901 and -5.525 for the approximately 7 mg/kg and approximately 10 mg/kg comparisons. For QMG, the reported differences are -3.483 and -4.756. Their sponsor-reported p-values are each < 0.001. Interpretation of change direction depends on the scoring rules for each measure; the posted MG-ADL description explicitly states that negative change indicates improvement.

General statistical explanation: standard errors quantify uncertainty in an estimate, while standard deviations describe variability among individual observations. Confusing them can make a group mean appear more or less precise than it is. The standard errors reported alongside the least-squares means are not individual-level standard deviations. The recalculation uses approximate standard deviations derived from those standard errors and group sizes, so its standardized effect summaries rely on that approximation.

General statistical explanation: a p-value is calculated under a null hypothesis and a statistical model. It describes how incompatible the observed data are with that null model, given the model’s assumptions. It is not a measure of effect size, clinical value, or the chance that a result occurred by coincidence. The reported p-values should be read alongside the estimated differences and confidence intervals.

Multiplicity arises when several outcomes or comparisons are evaluated. Testing many hypotheses can increase the chance of at least one small p-value if no adjustment or prespecified hierarchy is used. The registry record lists 14 analyses, but a multiplicity strategy is not reported in the registry record. Therefore, the listed p-values should not be assumed to be adjusted for all outcomes and comparisons. The record also does not report whether interim analyses were planned or performed.

General statistical explanation: over-dispersion occurs when data vary more than a statistical model expects. A model that does not account for extra variability can understate uncertainty. The primary endpoint is continuous and the record does not describe an over-dispersion assessment; there is no basis in the registry record to claim that over-dispersion was present or how it was handled. For a binary outcome, the uncertainty of a proportion or odds ratio likewise depends on the model and assumptions used.

The sample-size calculations are planning illustrations derived from observed effects, using a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. They examine effects set at stated fractions of the observed effect. They do not document the original design assumptions for the completed trial, which are not reported in the registry record. These planning figures are not observed power.

11. Confidence Intervals and Precision

The sponsor-reported primary difference for approximately 7 mg/kg versus placebo is -2.586, with a 95% confidence interval from -4.091 to -1.249. For approximately 10 mg/kg versus placebo, it is -2.619, with a 95% confidence interval from -3.994 to -1.163. Both intervals lie on the negative side of zero, and both reported p-values are < 0.001. In the scale’s stated direction, these intervals are consistent with a greater average reduction in MG-ADL score in each rozanolixizumab group than in placebo.

The intervals also show that the estimated contrast is not known with exact precision. They include a range of plausible values under the sponsor’s analysis method and assumptions. The interval for each comparison is useful alongside the point estimate: the estimate gives the central contrast, while the interval conveys uncertainty around it. The group standard errors are 0.488 for placebo, 0.486 for approximately 7 mg/kg, and 0.494 for approximately 10 mg/kg.

The recalculated intervals differ slightly from the sponsor-reported intervals. For approximately 7 mg/kg, the recalculated interval is -3.936 to -1.236; for approximately 10 mg/kg, it is -3.98 to -1.258. The recalculated point estimates match the sponsor’s reported point estimates, but the recalculated uncertainty is based on summary-statistic calculations and should not be mistaken for the trial’s fitted analysis.

Confidence intervals do not describe the range of responses among participants, and they do not account for questions not represented in the analysis, such as missing-data handling or multiplicity. The registry record does not report the primary analysis’s full modeling details or missing-data assumptions. Precision should therefore be understood as precision under the reported analysis, rather than certainty about every possible interpretation of the endpoint.

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 planning figures are participants per arm for 80% power and participants per arm for 90% power. They are not total sample sizes or total event counts. The calculations use the stated planning assumptions: a two-sided significance level of 0.05, equal allocation between the two arms, a normal approximation, and an effect to detect set at the stated fraction of the observed effect. The original trial’s own design assumptions and sample-size justification are not reported in the registry record.

In plain language, the 80% power column gives the estimated number of participants needed in each of two equally sized groups for a planned analysis to have 80% power under those assumptions. The 90% power column gives the corresponding number for 90% power. Power is the probability of detecting an effect of the specified size under a particular statistical model and assumptions, if that effect is present. These are planning illustrations, not observed power.

The rows for 100%, 75%, and 50% of the observed effect ask how planning requirements change if the effect of interest is set to the full observed estimate or to smaller fractions of it. Smaller effects are harder to distinguish from random variation, so they generally require more participants for the same target power. The figures should be read as conditional planning scenarios, not as evidence about what the trial originally planned or whether its design was adequate.

These calculations are based on the observed effect from this trial, so they may be optimistic if that estimate is larger than the effect expected in a future study. They also do not substitute for a design calculation using a justified target effect, expected variability, missing-data assumptions, and the intended analysis. The registry record does not report those original design details.

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 = -2.586; pooled standard deviation σ = 3.972.
2. n = 2 × (1.96 + 0.842)² × 3.972² / -2.586² = 37.0, rounded up to 38 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 <- -2.586                      # observed difference in means
sigma <- 3.97161071407082                     # 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(3971422)
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 = 38 (80% power), 50 (90% power)
 75% of observed effect: n per arm = 66 (80% power), 89 (90% power)
 50% of observed effect: n per arm = 149 (80% power), 199 (90% power)
power.t.test: 38 
Simulated power at n = 38 per arm: 0.794 (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

For the primary MG-ADL endpoint, the sponsor reports a least-squares mean difference of -2.586 for approximately 7 mg/kg versus placebo and -2.619 for approximately 10 mg/kg versus placebo. Both comparisons have reported p-values of < 0.001 and 95% confidence intervals that do not cross zero. The sponsor names the Lehmacher and Wassmer method for these continuous-outcome comparisons.

The unadjusted recalculations from the posted group summaries produce the same point estimates, with slightly different interval limits. The recalculations are not a replacement for the sponsor’s analysis: they use summary statistics and approximate variability, while the sponsor’s reported comparison reflects the analysis method named in the record. The registry discrepancy indicator is false, so there is no flagged disagreement that would change the reading of the posted results.

The same distinction applies to secondary outcomes. For MG-ADL response, the sponsor reports odds ratios from a Wald test, whereas the recalculated summaries include risk differences. These are different effect measures and are not expected to have numerically identical values. For continuous MG-C and QMG outcomes, the sponsor-reported least-squares mean differences are close to the corresponding unadjusted recalculated contrasts. Where the sponsor reports the trial analysis, those results remain the primary trial estimates; recalculations serve as supplementary checks.

15. Multiplicity, Missing Data and Analysis Populations

The record lists 9 outcome measures and 14 analyses, including a primary endpoint and multiple secondary outcomes and comparisons. When researchers evaluate several outcomes or make several group comparisons, multiplicity becomes relevant: the more hypotheses examined, the greater the opportunity for apparently unusual results by chance. The registry record does not report a multiplicity adjustment, a gatekeeping procedure, or a hierarchy for interpreting secondary tests. The reported secondary p-values should therefore be read as individual results, not assumed to control the overall false-positive risk across all analyses.

The primary outcome is clearly identified as change in MG-ADL score from baseline to Day 43. The record reports sponsor analyses for both dosage comparisons and labels them superiority analyses. It does not report whether the primary comparisons were part of a formally ordered testing sequence or whether the p-values account for the two comparisons. No interim analysis information is reported in the registry record.

The primary analysis population is described as the Randomized Set, with participants analyzed according to assigned treatment. This is an intention-to-treat-style assignment principle, but the record’s stated definition should be used rather than assuming additional details about inclusion or exclusion. The safety summaries have different group counts from the primary outcome summaries, and the registry record does not explain those differences in the data.

Missing-data methods are not reported in the registry record. The record does not state how missing Day 43 measurements were handled, whether imputation was used, or what sensitivity analyses were performed. Nor does it report the full model specification, covariates, or assumptions about missingness. These details can affect the interpretation of an estimated treatment contrast, so the posted estimates should be understood in light of the methods that are actually described, without inferring unreported procedures.

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 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: myasthenia gravis crisis (3.0%), myasthenia gravis (2.9%), gastritis (1.6%), vomiting (1.6%), arthralgia (1.6%).

The safety summary reports participants affected by serious adverse events in groups with 67 participants at risk for placebo, 64 for approximately 7 mg/kg, and 69 for approximately 10 mg/kg. The corresponding numbers affected are 6, 5, and 7. The record reports zero deaths in each of these groups. These are descriptive counts; by themselves they do not establish that the groups had equivalent safety profiles or explain the nature, timing, or duration of each event.

The record also reports participants with other, non-serious treatment-emergent adverse events: 26 in the placebo group, 40 in the approximately 7 mg/kg group, and 42 in the approximately 10 mg/kg group. The listed serious-event terms include myasthenia gravis crisis at 3.0%, myasthenia gravis at 2.9%, and gastritis, vomiting, arthralgia, and seizure at 1.6% each. The registry record does not provide arm-specific counts for these listed terms in the registry record.

Serious adverse events and other treatment-emergent adverse events are different summaries and should not be combined. Participant counts also do not indicate how many total episodes occurred. Safety interpretation requires context about event definitions, exposure, timing, severity, and attribution; those details are not reported in the registry record. These figures describe what was posted and should not be used alone to make a treatment decision.

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

The posted results show lower average MG-ADL scores at Day 43 relative to baseline in all three groups, with larger mean decreases in both rozanolixizumab groups than in placebo. The sponsor-reported group differences are negative, their 95% confidence intervals lie below zero, and both p-values are < 0.001. The available secondary numerical comparisons show a broadly consistent pattern across MG-ADL response, MG-C, and QMG, while their interpretation is tempered by multiple outcomes and the absence of a reported multiplicity strategy.

These results describe average differences in the randomized study population and at the stated time point. They do not mean that every participant experienced the same change. A group mean can coexist with substantial variation in individual outcomes, and the confidence interval for a mean difference is not a prediction interval for individual participants. The registry record does not report a threshold for a clinically meaningful MG-ADL change, so statistical evidence should not be equated automatically with a judgment about clinical importance.

The result is specific to the registered trial, its participants, its treatment groups, its endpoint, and its analysis. It does not establish the duration of any observed difference beyond Day 43, and the record does not provide later primary-outcome results. It also does not establish results in populations that were not represented in the randomized set. The posted outcomes do not answer every question about symptoms, functioning, safety, or treatment experience.

Finally, a small p-value is not proof, and a confidence interval is not a guarantee. They are statistical summaries that depend on the analysis and its assumptions. Missing-data handling, the complete statistical model, and multiplicity procedures are not reported in the registry record. The trial’s estimates should therefore be read as evidence about the measured comparisons, with the stated limits on precision and scope, rather than as a comprehensive account of all possible effects or outcomes.

18. Why This Trial Matters Statistically

This trial offers a useful example of how a continuous patient-reported outcome can be analyzed alongside binary and other continuous secondary measures. The MG-ADL endpoint has a defined scale and an explicitly stated direction: negative change indicates improvement, and higher scores indicate greater disability. That makes the sign of the reported contrast interpretable, while still requiring attention to the magnitude and uncertainty of the estimate.

The study has three randomized groups, allowing each dosage group to be compared with a shared placebo group. The posted group means, standard errors, sponsor-reported contrasts, confidence intervals, and p-values demonstrate how these summaries answer different questions. A point estimate describes the observed average contrast, a confidence interval conveys precision under the analysis, and a p-value evaluates compatibility with a null model. None alone describes the full clinical or statistical picture.

The record also illustrates why recalculations from summary statistics should be kept distinct from the sponsor’s analysis. The primary recalculated point estimates match the sponsor’s, while their interval limits differ. This is a reminder that reproducing a point contrast does not reproduce every modeling choice or uncertainty calculation. The sponsor’s analysis remains the principal trial result, and recalculations are a supplementary statistical check.

Finally, multiple outcomes and comparisons make multiplicity and transparency important. The registry identifies a primary endpoint and lists numerous secondary analyses, yet does not report a multiplicity strategy or detailed missing-data methods. Those omissions limit how strongly secondary p-values can be combined into an overall conclusion. Reading estimates together with their design, population, endpoint timing, and reporting limits is central to sound interpretation.

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 record does not report the full statistical analysis plan, the original sample-size justification, detailed randomization procedures, or whether interim analyses occurred. It also does not describe the complete model used for the primary least-squares means, including any covariates or other adjustments. These gaps limit independent assessment of how the reported analyses were specified and implemented.

Missing-data handling is not described. The record does not explain whether all randomized participants contributed Day 43 measurements, how any missing values were addressed, or whether sensitivity analyses tested different assumptions. The safety group counts differ from the primary outcome group counts, and the reason is not reported in the registry record.

The primary endpoint is measured at Day 43, so the result does not by itself describe later outcomes or persistence. The MG-ADL score captures specified patient-reported symptoms and disability, but it is not a complete measure of every relevant outcome. The registry record does not provide a clinically meaningful-change threshold, which limits interpretation of the estimate’s practical importance.

The study includes several secondary outcomes and comparisons. A multiplicity adjustment or testing hierarchy is not reported in the registry record, so unadjusted secondary findings should be interpreted cautiously. The available summary data also do not provide participant-level distributions, making it impossible to assess variation, outliers, or the shape of the outcome distribution from these registry entries alone. Recalculated intervals and standardized effects rely on summary-statistic approximations and are not substitutes for the trial’s full analysis.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: A Study to Test Efficacy and Safety of Rozanolixizumab in Adult Patients With Generalized Myasthenia Gravis — NCT03971422.
• Registry record last updated 2026-08-20; status last verified 2026-08.
• Results first posted 2023-08-21.
• 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

NCT03971422 is a completed phase 3 randomized, parallel-group, placebo-controlled study of rozanolixizumab in adults with generalized myasthenia gravis. Listed enrollment was 200. The primary endpoint was change from baseline to Day 43 in MG-ADL score, analyzed in the Randomized Set according to assigned group.

The sponsor-reported MG-ADL differences versus placebo were -2.586 for approximately 7 mg/kg and -2.619 for approximately 10 mg/kg; both had p < 0.001. The record also includes secondary outcome and adverse-event summaries. Details of the original sample-size justification, multiplicity procedures, and missing-data methods are not reported in the registry record.