Study to Evaluate the Efficacy and Safety of APL-2 in Patients With Paroxysmal Nocturnal Hemoglobinuria (PNH): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Study to Evaluate the Efficacy and Safety of APL-2 in Patients With Paroxysmal Nocturnal Hemoglobinuria (PNH) (NCT03500549), a randomized phase 3 study. The primary outcome is Least Squares (LS) Mean Change From Baseline to Week 16 in Hemoglobin (Hb) Level During the RCP.

2. Clinical Question

3. Trial Design

This was a Phase III, randomized, multicenter, open-label, active-comparator-controlled study with a crossover model. The registry lists an enrollment of 80 and a completed status. The recorded start date was 2018-06-14, and the primary completion date was 2019-11-14. These record details describe the trial's registration and conduct; they do not by themselves establish the completeness or quality of follow-up.

The posted treatment groups are sequences rather than simple labels for a single treatment throughout the study: Pegcetacoplan to Pegcetacoplan and Eculizumab to Pegcetacoplan. The primary endpoint and the three secondary outcomes with numerical results compare these groups during the RCP. The registry record does not describe the timing of every treatment period, the sequence allocation procedure, or how any period or carryover effects were addressed. The abbreviation RCP is not expanded in the record.

The study was open-label, with masking listed as none. That means participants and study personnel were not masked according to the registry entry. Open-label conduct can matter when outcomes involve treatment decisions, reporting, or assessment; Hb is a laboratory measure, but the record does not describe safeguards or laboratory blinding. The registry also does not report the original sample-size justification, the planned interim analyses, or the full statistical analysis plan. Those details should not be inferred from the posted enrollment or from retrospective planning illustrations.

The pegcetacoplan regimen is recorded as 1080 mg administered subcutaneously twice-weekly or every three days. Eculizumab is described as a complement (C5) inhibitor. The registry lists Soliris among interventions, while the treatment sequence uses the name eculizumab; it does not clarify that naming difference. The intention-to-treat set included all randomized subjects, according to the primary endpoint description.

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: The treatment period of the study consisted of 3 parts: a 4-week run-in period, a 16-week randomized controlled period (RCP), and a 32-week open-label period. Of the 102 subjects screened, 80 subjects met all the inclusion criteria and none of the exclusion criteria and entered the run-in period. Randomization was stratified by the number of packed red blood cell (PRBC) transfusions within the 12 months prior to Day -28 and platelet count at screening.

Run-in Period (Day -28 to ≤Day 1)

RCP (Day 1 - Week 16)

Reasons for not completing this period:

Open-label Period (Week 17 to Week 48)

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The ITT set included all randomized subjects.

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: The intent-to-treat (ITT) set included all randomized subjects.

6. Primary Endpoint

The registry lists the primary outcome as Least Squares (LS) Mean Change From Baseline to Week 16 in Hemoglobin (Hb) Level During the RCP, measured over Baseline and Week 16.

The primary endpoint was the least-squares mean change in Hb level from baseline to Week 16 during the RCP, measured in grams per deciliter (g/dL). Hb is a continuous laboratory measure. Change from baseline focuses the comparison on how the measured level moved over the specified period rather than only on the Week 16 value. A least-squares mean is a model-based estimate of an average; it can account for the analysis model's structure rather than being simply the arithmetic average of observed changes.

The record defines baseline as the average of measurements recorded before the first pegcetacoplan dose, including local and central laboratory values collected during screening. The analysis excluded data before the RCP and censored data for transfusions. These rules affect which observations contribute to the endpoint and how the treatment-period measurements are interpreted. The primary population was the intention-to-treat set, described as all randomized subjects.

Hb is used as an indicator of the measured blood hemoglobin level, and its repeated measurement can support comparison of change over time. The endpoint definition does not make the value a complete account of PNH symptoms, quality of life, or other clinical outcomes. The registry record does not provide the full model specification or explain all the assumptions used to handle repeated observations. The endpoint should therefore be read as a defined laboratory outcome, not as a stand-alone summary of every potential benefit or risk.

7. Results Posted in the Registry

• Eculizumab to Pegcetacoplan: mean -1.47 versus 2.37 on control; difference -3.84 (95% CI -5.33 to -2.35), standardized effect d = -1.15; 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 Subjects Who Did Not Require a Transfusion (Transfusion Avoidance) During the RCP

• 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:
  • Eculizumab to Pegcetacoplan: 15.4% (6/39) versus 85.4% (35/41) on control; risk difference -70.0 percentage points (95% CI -85.6 to -54.3), risk ratio 0.18 (0.09–0.38); unadjusted two-proportion z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically lower for Eculizumab to Pegcetacoplan than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

LS Mean Change From Baseline to Week 16 in Absolute Reticulocyte Count (ARC) During the RCP

• 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:
  • Eculizumab to Pegcetacoplan: mean 27.79 versus -135.82 on control; difference 163.61 (95% CI 137.06 to 190.16), standardized effect d = 2.74; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for Eculizumab to Pegcetacoplan than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

LS Mean Change From Baseline to Week 16 in Lactate Dehydrogenase (LDH) Level During the RCP

• 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:
  • Eculizumab to Pegcetacoplan: mean -10.12 versus -14.76 on control; difference 4.64 (95% CI -157.80 to 167.08), standardized effect d = 0.01; z-test p = 0.955.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Eculizumab to Pegcetacoplan and the control. The interval gives the range of effects that remain compatible with the data.

The posted secondary results cover outcomes that complement the primary Hb endpoint, including transfusion avoidance, ARC, LDH, fatigue, and bilirubin, as well as response and normalization measures. Among the outcomes with numerical results in the record, transfusion avoidance and ARC show clear separation between the posted group estimates, while the LDH comparison is imprecise and does not show a clear separation from zero in its recalculated interval. These measures describe distinct aspects of the trial outcomes; agreement or disagreement across them should not be treated as a single combined endpoint.

These secondary findings provide context for the primary Hb result because they concern transfusion use and other laboratory or functional measures. They do not establish that all outcomes moved in the same way. Numerical results for the remaining listed secondary outcomes are not reported in the registry record. The record lists 25 posted secondary outcomes, but the registry entries give detailed numerical results for only a subset; no missing values should be inferred from the outcome names.

Unadjusted secondary results carry less weight when many outcomes are examined. In general, conducting multiple tests increases the chance that at least one result will appear statistically notable by chance. The registry record does not report a multiplicity adjustment or a hierarchy governing secondary hypotheses. Consequently, secondary p-values should be treated as exploratory evidence rather than as independent confirmation of the primary result. Different endpoints may also have different scales, event frequencies, and analysis methods, limiting direct comparisons across them.

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

• Percentage of Subjects Who Did Not Require a Transfusion (Transfusion Avoidance) During the RCP
• LS Mean Change From Baseline to Week 16 in Absolute Reticulocyte Count (ARC) During the RCP
• LS Mean Change From Baseline to Week 16 in Lactate Dehydrogenase (LDH) Level During the RCP
• LS Mean Change From Baseline to Week 16 in Functional Assessment of Chronic Illness Therapy (FACIT) - Fatigue Scale Score During the RCP
• Percentage of Subjects Who Achieved a Hb Response in the Absence of Transfusions at Week 16
• Percentage of Subjects Who Achieved Reticulocyte Normalization in the Absence of Transfusions at Week 16
• Percentage of Subjects Who Achieved Hb Normalization in the Absence of Transfusions at Week 16
• LS Mean Change From Baseline to Week 16 in Indirect Bilirubin Level During the RCP

9. Statistical Interpretation

10. Statistical Methods Explained

For the continuous primary outcome, the registry reports least-squares mean changes and a sponsor analysis using a mixed model for repeated measures (MMRM). MMRM is a general approach for modelling repeated measurements from the same participants. It can use the pattern of measurements over time and account for within-participant correlation through a specified model. The precise covariance structure, fixed effects, and missing-data assumptions are not reported in the registry record, so they cannot be reconstructed from the posted summary.

What does a least-squares mean describe?

A least-squares mean is a model-based mean estimate. In general, model-based means can adjust for variables or repeated-measure structure included in the model. They are not necessarily identical to raw averages. The primary endpoint's posted values are described as least-squares mean changes, but the record does not provide enough model detail to identify every adjustment. The MMRM estimate is therefore best read as the sponsor's reported model-based comparison, not as a raw difference that can be recreated exactly from the arm summaries.

Why is the recalculation different in form?

The recalculated comparison is an unadjusted re-analysis from the posted summary statistics. It treats the two arm estimates as a simple contrast and uses approximate standard deviations inferred from the reported standard errors. That approximation is explicitly indicated for the continuous comparisons. It does not reproduce the repeated-measures model or its possible adjustments. The recalculated confidence interval and p-value are consequently illustrative checks on the posted summaries, not replacements for the sponsor's MMRM analysis.

For the primary endpoint, the recalculated contrast is -3.84, while the sponsor reports 3.84. The two estimates use opposite arm orderings: the recalculation states Eculizumab to Pegcetacoplan relative to Pegcetacoplan to Pegcetacoplan, whereas the sponsor's reported result is in the reverse direction. The signs therefore differ while the reported magnitude is 3.84. Their interval limits also differ because the methods are not the same. The record does not flag a discrepancy for this trial; this comparison is a reminder to check both estimand direction and analysis method before comparing estimates.

What do confidence intervals and p-values contribute?

In general, a confidence interval gives a range of values compatible with the data and the specified analysis, reflecting sampling uncertainty. Its width conveys precision: wider intervals permit a broader range of plausible effects. A p-value measures how unusual the observed data, or data more extreme, would be under a stated null model and its assumptions. It does not measure the probability that the null is true, the size of the effect, or the chance that an individual will experience an outcome. The registry's reported primary p-value is from the sponsor analysis; the recalculated unadjusted p-value comes from a different calculation.

The comparison estimates require attention to their direction. A positive contrast can mean something different from a negative one depending on which group is subtracted from which. The arm values and labels should be read alongside the contrast. In particular, the sponsor's primary LS mean difference is 3.84, while the simple comparison written in the reverse arm order is -3.84. Neither sign should be interpreted without its ordering.

Why can binary outcomes need different methods?

Transfusion avoidance is a binary outcome: each participant either did or did not require a transfusion during the specified period. A risk difference expresses a comparison between event proportions on an absolute scale. Risk ratios and odds ratios express relative comparisons, but their numerical interpretation differs. The sponsor's posted analysis for transfusion avoidance uses the Miettinen-Nurminen method for a risk difference and is labeled non-inferiority. The detailed margin and decision rule are not reported in the registry record. The unadjusted recalculation lists its own risk-difference estimate and interval; its method is not a substitute for the sponsor's analysis.

General statistical note: binary outcomes can be sparse or imbalanced, and the uncertainty of an estimate depends on the event counts as well as the denominators. It is useful to consider both absolute differences and relative measures, while keeping in mind that each answers a different question. No single scale is sufficient for every interpretation.

Why does over-dispersion matter?

General statistical note: over-dispersion occurs when observed variability is greater than a model's assumed variability. It can make standard errors too small if not handled, which in turn can make confidence intervals too narrow and p-values too optimistic. For count-like or proportion data, the variance may not follow a simple assumed pattern. The registry does not report an over-dispersion assessment or adjustment for these outcomes, so none should be presumed.

The trial used a crossover model, in which treatment sequence and period may matter. General statistical note: crossover data can include within-participant dependence, and treatment effects may be affected by period or carryover considerations. Appropriate analysis generally needs to reflect the design. The record does not explain how those issues were incorporated into the sponsor's models. The MMRM label alone does not provide the full specification.

Finally, the primary population is described as intention-to-treat, including all randomized subjects. The endpoint description also specifies censoring for transfusions and exclusion of data before the RCP. Those endpoint rules define the analysis data but do not, on their own, describe how other missing measurements were handled. That approach is not reported in the registry record.

11. Confidence Intervals and Precision

The sponsor's reported primary comparison is 3.84 g/dL, with a 95% confidence interval from 2.33 to 5.34. The interval lies above zero and is consistent with a positive difference in the sponsor's stated direction under the MMRM analysis. The recalculated unadjusted contrast is -3.84 g/dL, with a 95% confidence interval from -5.327 to -2.353, because its arm ordering is reversed. Both intervals indicate a separation from zero under their respective calculations, but their limits are not identical because the methods differ.

A confidence interval is not a range guaranteed to contain the true effect. In general, it describes the uncertainty associated with an estimate under the assumptions of the analysis. Its width is one way to assess precision: a narrower interval indicates less sampling uncertainty than a wider interval, all else equal. Precision is not the same as clinical importance, and an interval alone does not determine whether a difference matters to patients.

The LDH comparison illustrates why the point estimate should not be read alone. Its recalculated mean difference is 4.64 U/L, with a 95% confidence interval from -157.8 to 167.1 and an unadjusted p-value of 0.955. That interval includes values on both sides of zero and is wide, so the estimate is imprecise and compatible with a broad range of differences under the stated calculation. The sponsor's reported LDH interval is also broad, from -181.30 to 172.04.

For the binary transfusion-avoidance outcome, the recalculated risk-difference interval is from -0.856 to -0.543 in its stated comparison order. The sponsor reports a risk difference of 0.625 with a 95% confidence interval from 0.483 to 0.768 in the reverse direction. As with the primary endpoint, direction is essential to reading the sign. The ARC comparison also has a relatively bounded interval in the posted summaries, but its interpretation concerns a different laboratory measure. Precision should be considered outcome by outcome and method by method, rather than inferred from one result.

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 illustrations give participants per arm for continuous and binary comparisons. They are not totals across the study, and they are not total event counts. The stated assumptions are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. Each row uses a stated fraction of the observed effect as the effect to detect. The original trial's own design assumptions and sample-size justification are not reported in the registry record.

The 80% power column gives the illustrated number of participants per arm for a design that would have an 80% probability of detecting the specified effect under the stated assumptions, if that effect and assumptions were true. The 90% power column describes the corresponding planning target for a 90% probability. These are prospective planning calculations, not summaries of how likely the completed trial was to find its observed result.

Rows using a smaller fraction of the observed effect represent smaller target effects. Smaller effects are harder to distinguish from random variation, so the illustrations require more participants per arm. The figures are tied to the observed effect in this trial and should not be treated as a general sample-size recommendation. They also do not incorporate details absent from the listed assumptions, such as any particular crossover structure or planned adjustment. This is not observed power: it is a planning illustration based on the stated assumptions and effect fractions.

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 = -3.840; pooled standard deviation σ = 3.346.
2. n = 2 × (1.96 + 0.842)² × 3.346² / -3.840² = 11.9, rounded up to 12 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 <- -3.84                      # observed difference in means
sigma <- 3.34635498599711                     # 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(3500549)
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 = 12 (80% power), 16 (90% power)
 75% of observed effect: n per arm = 22 (80% power), 29 (90% power)
 50% of observed effect: n per arm = 48 (80% power), 64 (90% power)
power.t.test: 13 
Simulated power at n = 12 per arm: 0.767 (20000 replicates)

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

14. Analyses Reported by the Sponsor

The sponsor's primary analysis reports an MMRM LS mean difference of 3.84 g/dL, with a 95% confidence interval from 2.33 to 5.34 and p < 0.001. The simple recalculation from the posted arm summaries gives -3.84 g/dL, with an interval from -5.327 to -2.353 and p <0.001. The opposite signs reflect reverse ordering of the groups: the recalculation contrasts Eculizumab to Pegcetacoplan against Pegcetacoplan to Pegcetacoplan, while the sponsor reports the contrast in the other direction. The interval limits are not identical because the recalculation is unadjusted and uses approximate dispersion derived from the standard errors, whereas the sponsor uses MMRM.

The sponsor's analysis is the registry's formal model-based result. The recalculation is a transparent, simple comparison of posted summaries and should not be mistaken for a reproduction of the sponsor's model. The record does not flag a discrepancy. For transfusion avoidance, ARC, and LDH, the sponsor also reports model-based analyses with their own methods and labels. Those estimates should be interpreted on their own terms and in the direction defined by their group ordering.

15. Multiplicity, Missing Data and Analysis Populations

The registry record lists a primary Hb endpoint and a broad set of secondary outcomes. It does not report whether the analysis controlled the family-wise error rate, used a hierarchical testing sequence, adjusted secondary p-values, or designated outcomes as exploratory. Without such details, the effect of multiple testing on the interpretation of secondary results cannot be determined. An unadjusted p-value for an individual comparison does not account for the number of other outcomes examined.

The primary endpoint description identifies the intention-to-treat set as all randomized subjects and states that data before the RCP were excluded and transfusion observations were censored. These are important population and endpoint rules. However, methods for missing measurements, withdrawals, deviations from treatment, and incomplete follow-up are not reported in the registry record. The record also does not state whether sensitivity analyses assessed alternative missing-data assumptions.

The model is listed as crossover, but the statistical plan is not fully described. The registry does not report how treatment period, sequence, within-participant correlation, or possible carryover was handled. The sponsor's MMRM label is not enough to infer those details. Likewise, the record does not report the protocol's sample-size justification, interim analyses, or full analysis plan. These omissions limit evaluation of whether the posted analyses followed a prespecified strategy and how robust the estimates are to alternative reasonable choices.

The record includes sponsor analyses labelled superiority and non-inferiority for different outcomes, but it does not provide the non-inferiority margin or enough decision rules to independently assess that framework. This matters because non-inferiority interpretation depends on a prespecified margin and design assumptions, not only on a p-value or confidence interval. The registry entry is useful for identifying the outcomes and reported estimates, but it does not contain all the detail needed for a full audit of multiplicity, missing-data handling, or analysis populations beyond the stated ITT definition.

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: haemolysis (6.5%), anaemia (5.1%), gastroenteritis (2.6%), hyperthermia (2.6%), haemolytic anaemia (2.6%).

The adverse-event record separates the run-in period, open-label period, and RCP groups. During the run-in period, 4 serious adverse events were recorded among 80 at risk; during the open-label pegcetacoplan period, 18 were recorded among 77 at risk. In the RCP, 5 serious adverse events were recorded among 39 at risk in the eculizumab group, and 7 among 41 at risk in the pegcetacoplan group. These are counts reported for different periods and group denominators. They should not be added across periods or treated as if every period represented a separate set of participants.

The record lists one death during the open-label period and none in the other listed periods. It also lists serious adverse-event terms including haemolysis at 6.5%, anaemia at 5.1%, gastroenteritis at 2.6%, hyperthermia at 2.6%, haemolytic anaemia at 2.6%, and abdominal pain at 2.6%. The registry does not establish that an event was caused by a treatment. Serious-event counts and event terms describe reported safety observations; differences across periods may reflect exposure time, group size, and reporting context, details not fully described in the record.

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

The primary result describes a difference between the posted group estimates for change in Hb from baseline to Week 16 during the RCP. It is a comparison of group-level, model-based averages under the trial's endpoint definition, not a prediction of an individual participant's experience. The reported analysis is statistically separated from zero in its stated direction, but statistical evidence alone does not establish the size of clinical importance. The registry record does not provide a threshold for a clinically meaningful Hb change.

The result applies to the trial's enrolled PNH population, treatment sequences, observation period, and transfusion-censoring rule. It does not establish what would happen in other populations, with different treatment schedules, or over a different duration. The record's open-label design and incomplete reporting of model specifications, missing-data handling, and multiplicity procedures also constrain broader interpretation.

Secondary outcomes offer additional perspectives but are not interchangeable with Hb change. The reported separation in transfusion avoidance and ARC, alongside the imprecise LDH comparison, illustrates that different measures can produce different degrees of precision and statistical evidence. Numerical results for several listed secondary measures are not reported in the registry record. Nor does the record provide enough information to assess a unified benefit-risk balance from all outcomes.

The safety figures are reported event counts and event terms by period, not proof of causation. They do not alone establish comparative safety, particularly because the periods have different at-risk denominators and the record does not provide detailed exposure time or adjudication information. The trial statistics therefore describe specific recorded comparisons and observations. They do not amount to medical advice or a complete judgment about treatment choice, effectiveness, or safety in an individual case.

18. Why This Trial Matters Statistically

This trial illustrates why statistical interpretation begins with the estimand: what outcome is compared, over what period, in which population, and under what rules. Here, the primary measure is change in Hb from baseline to Week 16 during the RCP, with data before the RCP excluded and transfusion observations censored. Without those details, the numerical contrast could be misread as a general change over the whole study or as an uncensored treatment effect.

It also demonstrates the importance of comparison direction. The primary sponsor estimate is 3.84, while the simple recalculation is -3.84 because the groups are ordered oppositely. Recognizing this prevents a sign difference from being mistaken for a substantive disagreement. At the same time, the intervals differ because a model-based repeated-measures analysis and a simple unadjusted recalculation answer the comparison using different methods.

The crossover model raises questions about period and within-participant structure that cannot be answered from the registry entry alone. The record's limited statistical-plan detail is therefore not a minor technical point: it affects the ability to evaluate how the design was represented in the analysis. Finally, the variety of secondary measures shows why a statistically notable secondary result should not be treated as independent confirmation without knowing the multiplicity strategy. This case is a practical example of reading estimates together with their definitions, methods, precision, and reporting limits.

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.

Several trial-specific limitations follow from the registry record. The study was open-label, with masking listed as none. This may affect treatment-related decisions, participant reporting, or assessment, although Hb is a laboratory endpoint. The registry does not describe whether laboratory assessment was blinded or what safeguards were used.

The crossover model and sequence labels are reported, but the record does not give the schedule of all periods, explain the RCP abbreviation, or specify how period effects, sequence effects, and carryover were considered. Those details are important for evaluating a crossover analysis. The sponsor reports MMRM for the primary comparison, but the model specification is not available in the record.

The endpoint definition includes exclusion of data before the RCP and censoring for transfusions. The registry does not explain how other missing measurements or discontinuations were handled, or report sensitivity analyses. The full statistical analysis plan, original sample-size justification, and interim analysis details are also not reported in the registry record.

Multiplicity is another limitation: many secondary outcomes are listed, but the record does not state whether secondary tests were adjusted or sequenced. Numerical results are not available for all listed outcomes. The non-inferiority margins behind some sponsor-labelled analyses are not reported, limiting independent assessment of those claims. Finally, adverse-event counts are divided among study periods with different at-risk denominators, and the record does not provide enough context about exposure time or causality to support a simple comparative safety conclusion.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Study to Evaluate the Efficacy and Safety of APL-2 in Patients With Paroxysmal Nocturnal Hemoglobinuria (PNH) — NCT03500549.
• Registry record last updated 2022-03-25; status last verified 2022-03.
• Results first posted 2022-03-25.
• 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

NCT03500549 is a completed Phase III randomized, multicenter, open-label, active-comparator-controlled crossover study in PNH. The registry lists enrollment of 80, a start date of 2018-06-14, and a primary completion date of 2019-11-14. The treatment sequences posted for the primary comparison are Pegcetacoplan to Pegcetacoplan and Eculizumab to Pegcetacoplan.

The primary endpoint was least-squares mean change in Hb from baseline to Week 16 during the RCP. The sponsor reported an MMRM LS mean difference of 3.84 g/dL, with a 95% confidence interval from 2.33 to 5.34 and p < 0.001. The simple unadjusted recalculation uses the opposite group order and gives -3.84 g/dL. Secondary results include transfusion avoidance, ARC, and LDH; other statistical-plan and outcome details are not reported in the registry record.