Safety and Efficacy of Efavaleukin Alfa in Subjects With Steroid Refractory Chronic Graft Versus Host Disease: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Safety and Efficacy of Efavaleukin Alfa in Subjects With Steroid Refractory Chronic Graft Versus Host Disease (NCT03422627), a non-randomized phase 1, phase 2 study. The primary outcome is Phase 1b: Number of Participants Who Experienced a Dose-limiting Toxicity (DLT).

2. Clinical Question

The study therefore addresses a safety and tolerability question in Phase 1b, rather than a controlled estimate of treatment efficacy. The planned Phase 2 efficacy question remains unanswered by this study record. Although 32 participants were enrolled overall, that enrollment figure does not establish how many were evaluable for the DLT outcome or what their outcomes were. The registry record does not report the primary endpoint result or the detailed analysis population counts.

3. Trial Design

NCT03422627 was a Phase 1b/2, open-label, non-randomized, single-group study sponsored by Amgen. It enrolled 32 participants with chronic graft-versus-host disease. The study started on 2018-04-27, had a primary completion date of 2022-10-13, and is listed as terminated.

Phase 1b was designed to evaluate the safety and tolerability of multiple ascending doses of efavaleukin alfa. The registry describes five cohorts, each intended to include between 3 and 6 participants. Efavaleukin alfa was administered subcutaneously either every week or every 2 weeks, with protocol-permitted background therapy, for 52 weeks. The registry description does not provide the complete dosing details for each cohort.

Phase 2 was planned as a single-arm, multicenter, open-label study. Participants were to receive the recommended Phase 2 dose with protocol-permitted background therapy for up to 52 weeks. Its efficacy objective was overall response rate at 16 weeks, assessed using the 2014 cGVHD NIH Consensus Criteria. The registry explicitly states that early termination meant the Phase 2 portion was not conducted.

Because treatment was not randomized and there was no separate control group, comparisons between dose cohorts are not controlled treatment comparisons. Cohorts may have differed in dose or dosing schedule, and the record does not report a design that would make them interchangeable. The registry record does not report the reason for termination, the intended allocation of participants among cohorts, or a statistical analysis plan. Those details should not be inferred from the enrollment or cohort figures.

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: Screening tests and procedures were performed up to 28 days preceding treatment. The study was planned to have a Phase 1b part and a Phase 2 part. The study was cancelled before enrolling participants into the Phase 2 part. Therefore, results presented only include data for the Phase 1b part.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Measured in the DLT analysis set, which included DLT-evaluable participants in the phase 1b portion of the study. To be evaluable for a DLT participants must have received at least 2 doses of efavaleukin alfa or have experienced a DLT within the DLT evaluation period. The DLT evaluation period was defined as 4 weeks after the first dose of the study drug.

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: Measured in the safety analysis set, which included all participants who received at least 1 dose of efavaleukin alfa.

6. Primary Endpoint

The registry lists the primary outcome as Phase 1b: Number of Participants Who Experienced a Dose-limiting Toxicity (DLT), measured over Up to 4 weeks after first dose of study drug administration.

The registered Phase 1b primary outcome is the number of participants who experienced a dose-limiting toxicity (DLT) up to 4 weeks after the first dose of study drug administration. A DLT was defined using specified severity and attribution criteria. These included certain severe non-hematological toxicities related to efavaleukin alfa, prolonged severe decreases in neutrophil or platelet counts when attributed to the drug, and certain serious constitutional events. The record also specifies that only the most severe infections considered related to efavaleukin alfa were reviewed to determine whether they qualified as DLTs.

This endpoint is used in dose-escalation research to identify toxicities that meet a prespecified severity definition during an early evaluation period. Counting participants who meet that definition can inform a safety and tolerability assessment for the dose levels studied. A DLT count is not a complete account of all adverse events: events that do not meet the DLT definition may still be clinically important, and events occurring outside the evaluation period are not captured by this endpoint.

The DLT population was described as participants evaluable in Phase 1b. To be evaluable, a participant had to receive at least 2 doses of efavaleukin alfa or experience a DLT within the evaluation period. The remainder of the population description is incomplete in the registry record. No primary endpoint count or result is reported in the available record, so the observed DLT frequency cannot be characterized from this information.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Statistical methods help define what a result can say, but the registry record does not report a statistical analysis plan for this study. The outcome description and trial design can still be interpreted in general terms, while distinguishing those general concepts from methods actually documented for this trial.

What does a DLT count measure?

A count endpoint records how many participants meet a defined event criterion. Here, the criterion is a dose-limiting toxicity during the stated evaluation period, with specific severity and attribution requirements. A participant contributes to the endpoint only if the event satisfies the definition and the participant is in the DLT-evaluable population. This is different from counting every adverse event, every serious adverse event, or every laboratory abnormality.

Using a defined period and event threshold makes the endpoint more consistent across participants. It also narrows its scope: a toxicity outside the period, or one that does not meet the specified severity or attribution criteria, is not a DLT under this definition. The DLT count therefore addresses a particular early safety question rather than summarizing all potential harms over the full planned treatment period.

Why does the analysis population matter?

The registry describes the DLT analysis set as evaluable participants in Phase 1b. A participant was evaluable after receiving at least 2 doses or experiencing a DLT within the evaluation period. This definition affects the denominator used to interpret any DLT count. The total enrollment cannot be assumed to equal the number evaluable, and the record does not give the final evaluable count or the primary outcome result.

In general, excluding participants who do not meet an evaluability definition can make a reported count answer a different question from an analysis of everyone enrolled. A complete report would clarify how many participants entered each analysis population, how exclusions were handled, and whether participants who stopped treatment early remained under observation for safety. Such details are not reported in the registry record.

What does an open-label, single-group design imply?

In an open-label study, participants and study personnel are not masked to treatment assignment. In a single-group study, there is no separate concurrent control group. This design can describe outcomes observed during treatment and can support early assessment of tolerability, but it does not separate treatment effects from other influences as a randomized comparison is intended to do.

The Phase 1b study used multiple ascending doses across cohorts. Cohort observations may help characterize experience at the doses and schedules used, but the registry does not describe a randomized allocation between cohorts. Differences in outcomes across cohorts should therefore be treated as descriptive, not as an unbiased estimate of the effect of changing dose. No adjustment methods for cohort comparisons are reported.

How should uncertainty be assessed?

In general, a confidence interval describes the range of values compatible with the observed data under the assumptions of the analysis. It communicates precision as well as the estimate itself. Small or limited datasets often produce wide intervals, meaning that the data are compatible with a broad range of underlying event rates. A count without its evaluable denominator gives even less information about precision.

The registry record does not report a confidence interval for the DLT endpoint or a statistical analysis for that outcome. Consequently, no estimate of its precision can be drawn from the endpoint description alone. The same limitation applies to the Phase 2 response objective, which was not conducted according to the record.

What do the planning assumptions represent?

The registry includes assumptions for sample-size planning illustrations based on an observed effect: a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. These are assumptions for hypothetical planning, not a description of the actual trial's design. The actual study was non-randomized and single-group, and the Phase 2 portion was not conducted. The illustrations are not observed power and do not demonstrate that the trial was adequately powered for a clinical question.

In general, sample-size planning depends on the outcome, expected effect, variability, allocation, significance level, and desired power. A planning exercise can show how those assumptions relate to a proposed study size, but it does not guarantee that a study will produce a precise estimate or detect an effect. Since the record does not report a prospective sample-size justification for this trial, the planning assumptions should not be attributed to the original protocol.

Why might over-dispersion matter?

In general, over-dispersion occurs when event variability is greater than a statistical model assumes. It can make standard errors too small if the model does not account for the extra variability, leading to intervals that are too narrow and tests that are too optimistic. This issue is relevant to some count-data models, but the primary outcome is described as a count of participants with an event, and the registry record does not report a model or an over-dispersion adjustment. No claim can therefore be made about whether such a method was used or needed in the study's analysis.

These concepts clarify why design, endpoint definition, analysis population, and uncertainty all matter. They do not replace the missing outcome result. The registry record does not report the analysis methods needed to reproduce or evaluate a primary endpoint estimate.

10. Exact Estimates for Each Arm

No two-arm comparison is shown for this trial, but each arm's result can still be summarized with exact interval estimates computed from the posted counts. The Clopper-Pearson interval is exact and conservative; the Wilson interval is a common closer-to-nominal alternative; the Bayesian credible interval uses a uniform Beta(1, 1) prior. For small groups the three can differ noticeably.

R code and output

arm    <- c("Cohort 1", "Cohort 2", "Cohort 3", "Cohort 4", "Cohort 5")
n      <- c(6, 7, 6, 6, 3)
events <- c(0, 0, 0, 0, 0)
for (i in seq_along(n)) {
  ci <- binom.test(events[i], n[i])$conf.int          # Clopper-Pearson exact interval
  cr <- qbeta(c(0.025, 0.975), 1 + events[i], 1 + n[i] - events[i])   # Bayesian, uniform Beta(1,1) prior
  wi <- (function(x, n) { p <- x / n; z <- qnorm(0.975); d <- 1 + z^2 / n
    c((p + z^2 / (2 * n) - z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d, (p + z^2 / (2 * n) + z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d) })(events[i], n[i])
  cat(sprintf("ARM|%s|%d|%d|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f\n", arm[i], n[i], events[i], 100 * events[i] / n[i],
              100 * ci[1], 100 * ci[2], 100 * wi[1], 100 * wi[2], 100 * cr[1], 100 * cr[2]))
}

ARM  Cohort 1  6  0  0.0  0.0  45.9  0.0  39.0  0.4  41.0
ARM  Cohort 2  7  0  0.0  0.0  41.0  0.0  35.4  0.3  36.9
ARM  Cohort 3  6  0  0.0  0.0  45.9  0.0  39.0  0.4  41.0
ARM  Cohort 4  6  0  0.0  0.0  45.9  0.0  39.0  0.4  41.0
ARM  Cohort 5  3  0  0.0  0.0  70.8  0.0  56.1  0.6  60.2

11. Multiplicity, Missing Data and Analysis Populations

The registry record lists 4 outcome measures, but it does not report a multiplicity procedure. Multiplicity refers to the additional risk of chance findings when several outcomes, time points, subgroups, or comparisons are tested. In general, a prespecified testing strategy can help distinguish a primary confirmatory question from supportive or exploratory analyses. Without a reported strategy, the handling of multiple statistical tests cannot be assessed. The number of listed outcome measures alone does not establish how many were formally tested.

The record does not report a statistical analysis plan, analysis methods, or interim analyses. It also lists no analysis records. The absence of an analysis record does not establish that no analysis was ever performed; it means that results and methods are not reported in the registry record available for this trial.

Missing-data handling is not reported in the registry record. The information does not explain how missing endpoint assessments, early discontinuations, or participants who were not DLT-evaluable were handled in an analysis. It is therefore not possible to assess whether missing outcomes were excluded, imputed, or treated under another rule. No missing-data assumptions should be inferred.

The DLT endpoint is tied to a DLT-evaluable population, defined in part by receipt of at least 2 doses or experiencing a DLT during the evaluation period. The full population description is incomplete, and the number of evaluable participants is not reported. The record also does not give a Phase 2 analysis population because Phase 2 was not conducted. These gaps prevent a complete assessment of denominators, missingness, and the relationship between enrollment and analyzed participants.

12. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: drug hypersensitivity (40.0%), pneumonia (28.6%), pseudomonal sepsis (20.0%), capillary leak syndrome (20.0%), cellulitis (16.7%).

The registry lists serious adverse-event figures separately for cohorts 1–5. For each cohort, it gives the number at risk and the number affected by serious events and other events, along with deaths. For example, cohort 1 lists 5 participants affected by serious events among 7 at risk and 6 affected by other events among 7 at risk; cohort 2 lists 3 affected by serious events among 7 at risk and 6 affected by other events among 7 at risk. These are cohort-level descriptive counts, not randomized comparisons.

Deaths are listed as 0 for cohort 1, 1 for cohort 2, 1 for cohort 3, 0 for cohort 4, and 0 for cohort 5. The registry record does not report whether the deaths were attributed to study drug, cGVHD, or another cause. They should not be described as treatment-related on the basis of these counts alone.

The registry also lists serious-event terms and percentages, including drug hypersensitivity at 40.0%, pneumonia at 28.6%, pseudomonal sepsis at 20.0%, capillary leak syndrome at 20.0%, cellulitis at 16.7%, and parainfluenzae virus infection at 16.7%. The available record does not fully describe the denominator or attribution for those percentages. Serious events are not necessarily DLTs: DLT status depends on the separate definition, evaluation period, severity, and relationship criteria.

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

The record documents a study intended to assess efavaleukin alfa in steroid-refractory cGVHD, with a Phase 1b focus on safety and tolerability and a planned Phase 2 efficacy evaluation. It also documents early termination and states that the Phase 2 portion was not conducted. These registry entries describe the study's scope and status, but they do not establish whether the treatment improved outcomes or whether it did not.

The primary Phase 1b endpoint was a count of participants experiencing a DLT during a defined period. The registry record does not report the result of that endpoint, the DLT-evaluable population size, or an associated estimate of uncertainty. The serious adverse-event counts provide some descriptive safety information by cohort, but they do not replace the primary endpoint result and do not establish causality.

The planned Phase 2 overall response rate at 16 weeks is not an observed study result because that part of the trial was not conducted. There is no basis in the registry record for estimating the response rate or comparing it with another treatment. The trial's single-group, non-randomized design would also limit causal interpretation even if descriptive outcomes were available.

Accordingly, the study record can be used to understand what investigators planned to evaluate, the broad design, and selected safety counts. It cannot answer the central Phase 1b primary outcome question from the reported information, and it cannot answer the planned Phase 2 efficacy question. Missing results should be treated as unreported rather than as evidence of either benefit or harm.

14. Why This Trial Matters Statistically

This trial is statistically important as an example of why a study's stated objective and its reported results must be kept distinct. Phase 1b was designed around a defined DLT outcome, while Phase 2 was intended to assess response. The registry reports that Phase 2 was not conducted, and it does not report the Phase 1b primary outcome result. A reader therefore needs to separate the intended questions from the evidence actually available in the record.

The design also makes comparator language especially important. A non-randomized, single-group dose-escalation study can describe treatment exposure and observed events, but it does not create a like-for-like control group. Cohort-by-cohort figures are useful descriptively, yet they cannot be treated as randomized evidence about the relative effects of different doses or schedules.

Safety endpoints depend on definitions and analysis populations. The DLT definition includes severity, timing, and attribution criteria, while the DLT analysis set includes an evaluability rule. Without the final denominator and event count, a meaningful estimate of the endpoint's frequency cannot be reconstructed. The listed serious adverse events offer a separate view of safety and should not be conflated with DLTs.

Finally, planning illustrations based on an observed effect are not a substitute for an original sample-size rationale or reported inferential analysis. Their assumptions do not describe the actual single-group trial, and they are not observed power. This distinction is broadly relevant when interpreting registry records: design assumptions, descriptive outcomes, and confirmatory statistical evidence answer different questions.

15. Limitations

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

The most consequential limitation is incomplete outcome reporting. The registry record names the Phase 1b primary DLT endpoint but does not report its result or the size of the DLT-evaluable population. It also states that the Phase 2 portion was not conducted, so its planned overall response rate cannot be evaluated as an observed outcome.

The single-group, open-label, non-randomized design limits comparisons. There is no separate control group, and dose cohorts should not be treated as randomized arms. The registry record does not report methods that would support adjusted comparisons among cohorts. The study's design therefore cannot establish a treatment effect relative to a comparator from the available information.

The overall enrollment of 32 does not resolve the missing denominator for the primary endpoint. The population description for DLT evaluation is truncated, and the record does not report how many participants met the evaluability criteria. It also does not report the handling of missing assessments, the reasons for termination, a statistical analysis plan, or a sample-size justification.

Serious adverse-event figures are descriptive and reported by cohort. The record does not fully specify the denominator and attribution for the listed event-term percentages, and it does not provide enough information to assign causality to the reported deaths. These limitations restrict interpretation of both efficacy and safety. They should not be resolved by inferring unreported results from enrollment, study status, or planning assumptions.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Safety and Efficacy of Efavaleukin Alfa in Subjects With Steroid Refractory Chronic Graft Versus Host Disease — NCT03422627.
• Registry record last updated 2023-11-22; status last verified 2023-10.
• Results first posted 2023-11-22.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

19. Record Summary

NCT03422627 was a Phase 1b/2, open-label, non-randomized, single-group study of efavaleukin alfa in chronic graft-versus-host disease. It enrolled 32 participants and is listed as terminated. The study started on 2018-04-27, with a primary completion date of 2022-10-13.

Phase 1b evaluated multiple ascending doses, with DLTs as the registered primary outcome. The planned Phase 2 objective was overall response rate at 16 weeks under the 2014 cGVHD NIH Consensus Criteria. The registry states that Phase 2 was not conducted after early termination. The Phase 1b primary outcome result and its evaluable population size are not reported in the registry record.

Cohort-level serious adverse-event counts and selected event terms are listed, but the available information does not establish causality for deaths or fully clarify the denominators for listed percentages. No analysis records are listed, and the registry record does not report the statistical analysis plan, missing-data methods, or multiplicity procedures.