Safety and Efficacy of Eculizumab in Refractory Generalized Myasthenia Gravis (REGAIN Study): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Safety and Efficacy of Eculizumab in Refractory Generalized Myasthenia Gravis (REGAIN Study) (NCT01997229), a randomized phase 3 study. The primary outcome is Myasthenia Gravis Activities of Daily Living Profile (MG-ADL): Change From Baseline in MG-ADL Total Score at Week 26 by Worst-Rank Analysis of Covariance (ANCOVA).

2. Clinical Question

The registry identifies this as a Phase 3 randomized study with a parallel-group design. Its official title describes a double-blind, placebo-controlled, multicenter study; the masking field is listed as quadruple. The posted primary outcome is not a simple comparison of raw MG-ADL changes. It is a rank-based analysis in which all 125 participants were ordered from best outcome, rank/score 1, to worst outcome, rank/score 125, and the comparison was made using ANCOVA.

The primary reported estimates are 56.6 for eculizumab and 68.3 for placebo. In this ranking, lower ranks correspond to better outcomes by the stated ordering. The estimated mean difference is -11.7, with a 95% confidence interval from -24.33 to 0.96 and p = 0.070 in the sponsor-reported analysis. These results describe the posted analysis; they do not establish the clinical importance of the difference or answer questions about outcomes beyond the stated endpoint and time point.

3. Trial Design

REGAIN (NCT01997229) is a completed Phase 3 randomized parallel-group study sponsored by Alexion Pharmaceuticals, Inc. The official title describes it as double-blind, placebo-controlled, and multicenter; the registry masking classification is quadruple. Enrollment was 125, with 62 participants in the eculizumab group and 63 in the placebo group. The registry lists refractory generalized myasthenia gravis as the condition under study.

The intervention groups were eculizumab and placebo. The registry describes an induction phase followed by a maintenance phase for both arms. Its eculizumab description specifies 3 vials, equivalent to 900 mg, weekly for 4 doses, followed by 4 vials, equivalent to 1200 mg, one week later for the fifth dose at Week 4. The description is truncated before the maintenance schedule is complete. The placebo description is also truncated, so the complete schedules are not reported in the registry record.

The study began in 2013-12, reached primary completion in 2016-02, and is listed as completed. The primary outcome timeframe is end of study at Week 26. One primary outcome measure and one analysis are posted. The registry record does not report the full protocol, detailed eligibility criteria, participant flow, allocation implementation, the roles included in the quadruple masking, or the original design assumptions for sample size. Those details cannot be inferred from the randomized, masked, parallel-group labels alone.

The posted primary analysis compares the groups on MG-ADL change from baseline through a worst-rank ANCOVA. The record defines the ranking across all 125 participants from best to worst. This matters for interpretation: the reported group estimates are rank-based analysis results, not direct summaries of the original MG-ADL score scale. The exact covariates and other model specifications 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.

Overall Study

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 Myasthenia Gravis Activities of Daily Living Profile (MG-ADL): Change From Baseline in MG-ADL Total Score at Week 26 by Worst-Rank Analysis of Covariance (ANCOVA), measured over End of study (Week 26).

The primary endpoint is change from baseline in the MG-ADL total score at Week 26. MG-ADL is the Myasthenia Gravis Activities of Daily Living Profile, a measure of daily-life activity in people with myasthenia gravis. The posted unit is scores on a scale. The registry does not include additional details about the instrument’s scoring rules, component items, or the clinical meaning of a particular score change, so these should not be inferred from the endpoint label alone.

The analysis used a worst-rank approach. The registry states that all 125 participants were ranked from best outcome, rank/score 1, to worst outcome, rank/score 125. This approach places participants on a common ordered scale for the analysis rather than comparing only their raw changes. In general, ranking can be useful when the analysis is designed to account for outcomes that are difficult to compare directly or when a prespecified ordering is used to represent the overall outcome. The particular rationale for using this method in this trial is not reported in the registry record.

The posted estimates, 56.6 for eculizumab and 68.3 for placebo, are least-squares means from the rank-based ANCOVA. Because smaller ranks represent better outcomes in the stated ordering, the lower estimate for eculizumab is directionally favorable within this rank analysis. However, the confidence interval for the estimated difference includes zero, and the result should be read as an estimate with uncertainty rather than as a definitive conclusion. The ranking scale should not be interpreted as a raw MG-ADL point scale.

7. Results Posted in the Registry

• Eculizumab: mean 56.60 versus 68.30 on control; difference -11.70 (95% CI -24.20 to 0.80), standardized effect d = -0.33; z-test p = 0.067.

8. Statistical Interpretation

9. Statistical Methods Explained

The posted primary analysis is a worst-rank analysis of covariance (ANCOVA). Its target is the difference between randomized groups in the Week 26 change-from-baseline outcome after the endpoint values have been transformed into ranks. The registry says that all 125 participants were ordered from best outcome at rank/score 1 to worst outcome at rank/score 125. Because a lower rank represents a better position in that ordering, lower group estimates point in a favorable direction for that group on this particular analysis scale.

What does ranking change?

In a direct analysis of a continuous score, the numerical distances between values contribute to the comparison. A rank analysis instead uses the ordering of observations. The rank is determined by relative position among the participants included in the ranking, and the size of a difference in original score units is not carried over as the same numerical distance between ranks. Thus, the posted estimates of 56.6 and 68.3 are rank-based least-squares means. They should not be read as MG-ADL score changes or as numbers of scale points.

The word “worst-rank” describes the stated ordering of outcomes, with the worst outcome assigned the highest rank. The record’s brief description does not specify additional rules for ties, missing outcomes, or how particular events affected a participant’s rank. Those details are not reported in the registry record. It is therefore possible to explain the overall best-to-worst ordering, but not to reconstruct each participant’s rank or the full ranking algorithm from the posted summary.

What does ANCOVA contribute?

ANCOVA is a regression-based method for comparing group means while accounting for selected covariates. In general, adjustment can improve precision or address imbalance in measured variables, depending on the design and model. The registry labels the analysis ANCOVA and reports a least-squares mean difference. It does not report the covariates, the model formula, or further implementation details. The label alone does not identify what was adjusted for, so no specific adjustment should be assumed.

A least-squares mean is a model-based group estimate, rather than necessarily the arithmetic average of the observed participant values. The reported estimates are 56.6 for eculizumab and 68.3 for placebo, and the reported net mean difference is -11.7. The description calls the parameter a mean difference, while the endpoint analysis is based on ranks. The interpretation should therefore remain on the ranked outcome scale.

How should the standard errors be read?

The registry lists standard errors of 4.53 for the eculizumab estimate and 4.49 for placebo. A standard error describes the estimated sampling uncertainty of an estimate: in general, it reflects how much an estimate would vary across repeated studies under the analysis assumptions. It is not the spread of individual participant outcomes. The registry does not provide standard deviations for the ranked endpoint, and the listed standard errors should not be treated as measures of participant-to-participant variability.

A confidence interval combines an estimate with its uncertainty to give a range of values compatible with the data and analysis assumptions at the stated confidence level. The sponsor reports a 95% interval from -24.33 to 0.96 for the group difference. Since this range includes zero, it does not rule out a zero difference at that confidence level. The interval’s width also indicates that the estimate is not highly precise: differences in either direction remain compatible with the reported analysis.

What does the p-value describe?

The sponsor’s p-value is 0.070. In general, a p-value is calculated under a specified null model, commonly one in which the group difference is zero, and describes the probability of data at least as incompatible with that model as the observed data. It is not the probability that the null hypothesis is true, and it does not measure the size or clinical importance of the group difference. The registry reports a superiority analysis but does not provide the complete protocol or statistical analysis plan. The interpretation of the p-value should therefore stay tied to the reported ANCOVA and its stated result.

The unadjusted recalculation uses the posted estimates and standard errors, with an approximate pooled standard deviation also reported. It yields a difference of -11.7, a 95% confidence interval from -24.2 to 0.801, and p = 0.067. Such a recalculation is useful as a check on the general scale and uncertainty, but it is not a reconstruction of the ANCOVA. It does not reproduce the model’s covariate adjustment or any details not available in the summary statistics. The sponsor-reported ANCOVA remains the trial’s reported analysis.

Why can over-dispersion matter?

In general, over-dispersion means that observed variability is greater than a model expects. If a model assumes less variability than is actually present, standard errors can be too small and confidence intervals too narrow; if it accounts for extra variability, uncertainty estimates may be more realistic. This is a general statistical consideration, not a finding reported for this trial. The posted primary analysis is ANCOVA on ranks, and the registry does not report an over-dispersion assessment or indicate that over-dispersion was detected.

Finally, statistical significance and clinical importance are distinct. A small p-value does not by itself establish a meaningful benefit, while a p-value above a chosen threshold does not prove that groups are identical. The estimated difference, confidence interval, endpoint construction, and limitations of the available analysis details all inform interpretation. For this trial, the interval spans zero and the original model specifications are not reported in the registry record.

10. Confidence Intervals and Precision

The sponsor-reported estimate is a mean difference of -11.7 on the ranked outcome scale. Its 95% confidence interval is -24.33 to 0.96. A confidence interval is a range generated by a statistical procedure that, under its assumptions, reflects uncertainty about the underlying group difference. It is not a range containing 95% of individual participant outcomes, nor does it mean there is a 95% probability that the true value lies in this particular interval.

The interval includes zero, the value corresponding to no group difference. It therefore does not exclude no difference at the stated confidence level. Its negative values are compatible with a more favorable average rank for eculizumab, given the ranking where 1 is best; its positive end allows a small difference in the opposite direction. The interval is broad enough that the estimate’s direction and size remain uncertain. The registry does not report a threshold for a clinically important difference, so the clinical meaning of the interval’s endpoints cannot be classified from the record alone.

The recalculated unadjusted interval is -24.2 to 0.801, compared with the sponsor’s interval of -24.33 to 0.96. Both intervals include zero, but the sponsor’s ANCOVA is the reported trial analysis. The recalculation uses posted summary statistics and is not a substitute for the full model. The standard errors listed for the group estimates are 4.53 and 4.49; they quantify uncertainty in those estimates, not individual variation.

Precision depends on the information available and the analysis assumptions. The registry record does not provide participant-level results, full model specifications, or missing-data details, so the reported interval cannot be independently reconstructed in full. The confidence interval is the clearest posted expression of uncertainty around the primary estimated difference.

11. 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 event counts. The planning illustration assumes a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. It uses the observed effect as a starting point and evaluates effects set at 100%, 75%, or 50% of that observed effect. These are planning assumptions for the illustration; the trial’s original design assumptions are not reported in the registry record.

For the 100% observed-effect scenario, the illustration gives 146 participants per arm for 80% power and 196 participants per arm for 90% power. For planning at 75% of the observed effect, it gives 260 participants per arm for 80% power and 347 participants per arm for 90% power. For planning at 50% of the observed effect, it gives 584 participants per arm for 80% power and 781 participants per arm for 90% power.

In general, 80% power means that if the assumed effect and statistical conditions were true, a study of that size would have an 80% chance of detecting an effect at the stated significance level. The 90% column describes the corresponding 90% chance under the same assumptions. These planning probabilities do not describe the probability that the treatment effect is real. Smaller effects require more participants to achieve the same target power, which is why the illustrated participant counts rise as the assumed fraction of the observed effect gets smaller.

This is a planning illustration, not observed power. It uses the observed effect from this trial as an input and does not establish that the original study was designed using these assumptions. Because the trial’s own design assumptions are not reported in the registry record, these figures should be read as hypothetical planning quantities rather than as a reconstruction of the original sample-size justification.

12. 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 = -11.700; pooled standard deviation σ = 35.654.
2. n = 2 × (1.96 + 0.842)² × 35.654² / -11.700² = 145.8, rounded up to 146 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 <- -11.7                      # observed difference in means
sigma <- 35.6536403231519                     # 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(1997229)
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 = 146 (80% power), 196 (90% power)
 75% of observed effect: n per arm = 260 (80% power), 347 (90% power)
 50% of observed effect: n per arm = 584 (80% power), 781 (90% power)
power.t.test: 146.7 
Simulated power at n = 146 per arm: 0.801 (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.

13. Analyses Reported by the Sponsor

The sponsor-reported primary analysis is ANCOVA, with a net mean difference of -11.7, a 95% confidence interval from -24.33 to 0.96, and p = 0.070. The posted recalculation is a simple unadjusted analysis of summary statistics. It gives the same mean difference, -11.7, with a 95% confidence interval from -24.2 to 0.801 and p = 0.067.

The estimates are close but not identical in their inferential results. This is expected when a summary-statistic recalculation is not the same as the trial’s ANCOVA. The registry does not report the model covariates or full specification, so the sponsor’s analysis cannot be reproduced from the posted summaries alone. The recalculation is best understood as a supplementary numerical check, not a replacement for the trial analysis.

The sponsor’s confidence interval includes zero, as does the recalculated interval. Both therefore reflect uncertainty that includes no group difference. The sponsor’s reported ANCOVA is the primary reference for the trial result. The record does not report the original statistical analysis plan, so no further claims about prespecified procedures or design decisions can be made.

14. Multiplicity, Missing Data and Analysis Populations

The registry posts one primary outcome measure and one analysis. It lists no secondary outcome analyses. This indicates that the public results record summarized for this study contains the primary analysis but does not provide secondary outcome results; it does not establish whether secondary outcomes were planned or measured. The protocol’s full outcome list is not reported in the registry record.

Multiplicity arises when several statistical hypotheses are tested. In general, conducting many tests can increase the chance of at least one apparently unusual result, so a statistical plan may specify how to address multiple comparisons. The registry record does not report a multiplicity strategy, whether any adjustment was planned, or whether the primary endpoint was the only confirmatory analysis. The single posted primary analysis should not be taken as proof that no other tests were conducted.

Missing-data handling is also not reported. The record does not say how participants without Week 26 measurements were handled, whether any observations were imputed, or what assumptions were used for incomplete outcomes. The primary analysis description ranks 125 total patients, but it does not provide participant-level follow-up information or a detailed account of missing observations. The enrolled group sizes are 62 for eculizumab and 63 for placebo; these counts alone do not establish how many participants contributed to every part of the analysis.

The analysis population is not identified in the registry record. It does not specify whether the primary analysis used an intention-to-treat population, a modified intention-to-treat population, a per-protocol population, or another defined set. Nor does it report exclusions from the randomized groups. These omissions limit interpretation of the connection between randomization, the participants analyzed, and the reported estimates.

The record also does not report the ANCOVA covariates, detailed model assumptions, or handling of ties in the rank procedure. These are relevant to reproducibility and to understanding how the reported estimate was obtained. The posted result can be interpreted as the sponsor’s reported rank-based ANCOVA comparison, while these unreported details remain limits on independent evaluation.

15. 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 (12.7%), pyrexia (3.2%), upper respiratory tract infection (3.2%), metastases to bone (1.6%), prostate cancer (1.6%).

The registry reports serious adverse events for 9 of 62 participants in the eculizumab group and 18 of 63 participants in the placebo group. It reports other adverse events for 52 of 62 participants in the eculizumab group and 53 of 63 participants in the placebo group. These are participant counts within the reported at-risk groups; the figures do not describe event severity beyond the registry’s serious versus other classification, event duration, or individual clinical course.

No deaths are reported in either group: 0 for eculizumab and 0 for placebo. The listed serious adverse-event terms include Myasthenia gravis at 12.7%, Pyrexia at 3.2%, Upper respiratory tract infection at 3.2%, Metastases to bone at 1.6%, Prostate cancer at 1.6%, and Lymphopenia at 1.6%. The registry’s listed terms and percentages are not assigned to a particular arm in the available summary, so they should not be interpreted as arm-specific comparisons.

These figures describe the safety entries posted in the registry. They do not by themselves establish causation or permit a complete comparison of risks. The registry record does not provide fuller event narratives, timing, exposure-adjusted rates, or information needed to assess whether a reported event was judged related to study treatment.

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

The primary result is an estimated difference between the randomized groups on a rank-based analysis of Week 26 MG-ADL change. The eculizumab estimate is 56.6 and the placebo estimate is 68.3; the estimated difference is -11.7. Because lower ranks represent better outcomes in the registry’s ordering, the point estimate favors eculizumab on that analysis scale. The sponsor-reported 95% confidence interval, -24.33 to 0.96, includes zero, and the p-value is 0.070.

These numbers do not prove that the groups are the same, and they do not establish a definitive difference. The point estimate is uncertain, as reflected by the confidence interval. Nor does a directionally favorable estimate alone show that any difference is large enough to matter to participants. The registry record does not report a clinically important threshold for the MG-ADL outcome, and the rank scale is not equivalent to raw MG-ADL score points.

The posted safety counts show serious adverse events in 9 of 62 participants receiving eculizumab and 18 of 63 receiving placebo, other adverse events in 52 of 62 and 53 of 63, and no deaths in either group. These are descriptive registry entries, not proof that treatment caused or prevented any event. The record does not supply the detail required for a comprehensive safety assessment.

Interpretation is also limited by unreported analysis details. The registry does not report the full ANCOVA specification, missing-data methods, analysis population, multiplicity procedures, or original sample-size assumptions. Only one primary outcome analysis is posted, and no secondary analyses are posted. The result should therefore be read as the reported primary statistical comparison at the stated endpoint and time point, with the uncertainty and reporting limitations described, rather than as a complete account of all possible benefits, risks, or outcomes.

17. Why This Trial Matters Statistically

This trial illustrates how the analysis scale affects the meaning of a treatment comparison. The endpoint is described as change from baseline in MG-ADL, but the primary analysis ranks all 125 participants from best to worst and reports least-squares means of those ranks. A difference of -11.7 is therefore not a difference of -11.7 MG-ADL points. Recognizing that distinction is essential when reading estimates and considering whether they have a clinical interpretation.

It also demonstrates why an estimated direction and a precise conclusion are not the same thing. The point estimate is in the direction of a more favorable average rank for eculizumab, but the sponsor-reported interval extends from -24.33 to 0.96 and includes zero. The p-value is 0.070. The estimate, interval, and analysis method should be considered together; none alone supplies a complete account of the result.

The comparison between the sponsor’s ANCOVA and a simple unadjusted recalculation highlights the difference between reproducing a reported analysis and checking summary statistics. The recalculated estimate is similar, but the inferential values are not identical. Because the registry does not report the model’s detailed covariates or specification, the posted ANCOVA cannot be fully recreated from the summary alone.

Finally, the registry record shows why statistical interpretation depends on reporting context. Details of missing data, analysis populations, multiplicity, and the original sample-size justification are not reported. The planning illustration can demonstrate how sample-size needs change under different assumed effects, but it cannot be treated as the trial’s design rationale or as evidence of observed power. Together, these points make transparent reporting of the endpoint construction, model, uncertainty, and analysis population important for evaluating clinical trial results.

18. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The primary endpoint is reported through a worst-rank ANCOVA, but the registry’s brief description does not provide the full ranking procedure. The handling of ties, incomplete outcomes, or specific types of unfavorable outcomes is not reported. As a result, the overall best-to-worst ordering is clear, but the precise construction of every rank cannot be independently evaluated from the registry record.

The ANCOVA covariates and model specification are not reported. This limits the ability to reproduce the sponsor’s estimates or determine what adjustments were made. The participant population included in the primary analysis is also not reported, and the registry does not provide a participant flow account linking the 125 enrolled participants to the analyzed observations. Missing-data methods are not reported.

The registry posts one primary analysis and no secondary outcome analyses. It does not report whether additional outcomes or subgroup analyses were planned, whether multiplicity was addressed, or what the original sample-size assumptions were. The planning illustration based on the observed effect does not answer those questions and should not be mistaken for the original design calculation.

The arm descriptions are incomplete. The registry includes induction details for both groups but truncates the maintenance-phase descriptions, so the full treatment schedules cannot be reconstructed from this record. The safety entries are also limited: they provide counts for serious and other adverse events, deaths, and selected serious event terms, but do not report event timing, exposure-adjusted rates, detailed narratives, or treatment-relatedness assessments.

Finally, the rank-based estimates should not be translated directly into raw MG-ADL points, and the record does not provide a clinically important threshold to interpret the size of the difference. These limitations affect reproducibility, clinical interpretation, and the scope of conclusions that can be drawn from the posted results.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Safety and Efficacy of Eculizumab in Refractory Generalized Myasthenia Gravis (REGAIN Study) — NCT01997229.
• Registry record last updated 2019-07-16; status last verified 2019-07.
• Results first posted 2018-02-07.
• 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.

22. Record Summary

NCT01997229, REGAIN, is a completed Phase 3 randomized parallel-group study of eculizumab versus placebo in refractory generalized myasthenia gravis. The registry lists 125 enrolled participants: 62 in the eculizumab group and 63 in the placebo group. The official title describes a double-blind, placebo-controlled, multicenter study, and the registry masking classification is quadruple.

The primary outcome is change from baseline in MG-ADL total score at Week 26, analyzed using worst-rank ANCOVA. The sponsor-reported mean difference is -11.7, with a 95% confidence interval from -24.33 to 0.96 and p = 0.070. The interval includes zero. The rank-based estimate is not a raw MG-ADL point difference.

Serious adverse events were reported for 9 of 62 eculizumab participants and 18 of 63 placebo participants; no deaths were reported. The registry does not report secondary analyses, full model details, missing-data methods, analysis populations, multiplicity procedures, or the original sample-size assumptions.