Methoxsalen and Extracorporeal Photopheresis (ECP) for the Treatment of Pediatric Participants With Steroid Refractory Acute Graft Versus Host Disease: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Methoxsalen and Extracorporeal Photopheresis (ECP) for the Treatment of Pediatric Participants With Steroid Refractory Acute Graft Versus Host Disease (NCT02524847), a na phase 3 study. The primary outcome is Number of Participants Achieving Overall Response (OR) Using the Modified International Bone Marrow Transplant Registry (IBMTR) Severity Index at Week 4.

2. Clinical Question

3. Trial Design

This was a Phase 3, multicenter, single-arm, open-label study of methoxsalen used with ECP in pediatric participants with steroid-refractory aGvHD. The record lists enrollment of 29. With no randomized allocation and no control arm, the design can describe outcomes among participants receiving the study regimen but cannot directly estimate a treatment contrast against a concurrent alternative.

The recorded study periods were screening, treatment, and follow-up. Methoxsalen was given at 20 µg/ml in conjunction with ECP three times per week for Weeks 1 to 4 and two times per week for Weeks 5 to 12. The primary outcome was assessed at 4 weeks. Secondary outcomes included response at Week 8 and Week 12, duration of response within 16 weeks, response according to modified Glucksberg criteria, cumulative daily steroid dose, organ-stage assessments, and adverse events.

The study start date is 2016-01-20, and the primary completion date is 2019-07-16. Its status is terminated. The reason for termination, whether the treatment schedule changed, and whether all planned follow-up was completed are not reported in the registry record. The record names the sponsor as Therakos, Inc., a Mallinckrodt Company. It does not describe the details of site conduct, participant flow, or the handling of treatment discontinuation.

Because the study was unmasked and single-arm, its interpretation differs from that of a blinded randomized comparison. Response assessments may be susceptible to knowledge of treatment, while changes over time cannot be separated from the treatment regimen using a concurrent control. These are design properties, not evidence that a particular assessment was biased or that any observed outcome was caused by the regimen.

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: Participants were recruited by multiple treatment centers in the United States and Europe

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: All participants

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: Safety Analysis Set included all participants who receive any amount of study drug.

6. Primary Endpoint

The registry lists the primary outcome as Number of Participants Achieving Overall Response (OR) Using the Modified International Bone Marrow Transplant Registry (IBMTR) Severity Index at Week 4, measured over 4 weeks.

The primary endpoint is the number of participants achieving overall response at Week 4 under the modified IBMTR Severity Index. The registry defines overall response as complete response (CR) plus partial response (PR). CR means complete resolution of all signs and symptoms of aGvHD in all evaluable organs without addition of next-line systemic treatment. PR means improvement of 1 stage in 1 or more aGvHD target organs, without progression in others and without addition of next-line systemic treatment.

This endpoint condenses a set of organ-level clinical assessments into a response classification. Reporting a count can describe how many participants met the stated response definition at the specified assessment time. The definition also makes the treatment-change condition part of classification: improvement alone is not enough if next-line systemic treatment was added. The record does not report the actual Week 4 count, the proportion responding, or an uncertainty interval. It also does not report the detailed assessment procedures used to determine evaluability across organs.

In general, a response endpoint can be useful for describing whether participants meet prespecified clinical criteria over a defined period. It does not, by itself, establish that treatment caused the response, particularly in a single-arm study without a concurrent comparator.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does the primary count represent?

The primary endpoint is a count of participants meeting the overall-response definition at Week 4. The definition combines complete response and partial response according to the modified IBMTR Severity Index. A count is a direct summary of how many people satisfy a classification, but its interpretation depends on knowing the population assessed and how missing or unevaluable assessments were handled. Although the primary population is labelled all participants, the registry record does not report the number with an evaluable Week 4 assessment or the rules used when an assessment was unavailable.

In general, a count can be accompanied by a proportion to put it in relation to an analysis denominator. That denominator must be defined rather than assumed: it might involve all enrolled participants or a specified subset, depending on the analysis plan. The registry record does not report a primary response count or denominator for a calculated response proportion. No response proportion should therefore be inferred from enrollment alone.

How should uncertainty be understood?

In general, a confidence interval describes the range of effect estimates reasonably compatible with the data under a statistical model. It helps readers judge precision: a point estimate alone does not show how much uncertainty surrounds it. For a binary response outcome, precision depends on both the number meeting the definition and the number assessed. The registry record does not provide a primary confidence interval or enough reported outcome information to derive one without making assumptions. No interval should be treated as a substitute for the sponsor’s analysis.

A p-value, in general, evaluates how compatible the data are with a specified null model; it is not the probability that a treatment works, nor does it measure the size or clinical importance of an effect. A p-value for the primary endpoint is not reported in the registry record. The record also does not report a statistical model, a null hypothesis, or a test used for the primary outcome. Without those details, there is no basis to reconstruct or interpret a primary significance test.

What does a single-arm design allow?

A single-arm study observes outcomes in participants receiving the regimen, without a concurrent control group. This can describe the observed course in the enrolled group, but does not estimate what would have happened to comparable participants under another intervention. In general, a before-and-after comparison can also be influenced by changes in disease status, concurrent care, assessment practices, and participant characteristics over time. A randomized control group can help distinguish treatment contrasts from such influences; this study’s registry record does not list one.

Open-label means masking was not used. In general, knowledge of assigned treatment can influence reporting, clinical management, or outcome assessment, depending on the endpoint and procedures. The modified IBMTR response classification uses assessments of signs and symptoms across evaluable organs, so clear and consistently applied assessment rules matter. The registry record does not report whether independent reviewers were used, whether assessments were blinded, or how disagreements were resolved. These details should not be assumed.

How do the planning calculations relate to this study?

Sample-size planning asks how many participants might be needed under stated assumptions to detect a target effect with specified statistical properties. Such a calculation is prospective or illustrative; it is not a measure of the probability that the completed study produced a statistically significant result. The planning assumptions associated with the posted illustrations specify a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. They also define target effects as stated fractions of an observed effect.

Those assumptions describe a hypothetical two-arm setting, whereas this trial is recorded as single-arm. The illustrations are therefore planning scenarios, not a description of the actual design or an observed power estimate. The record does not report the trial’s own sample-size justification, assumed target effect, or statistical plan. The primary effect needed as an input to an observed-effect-based illustration is also not reported in the registry record. The illustrations cannot establish whether enrollment was adequate for the trial’s original objectives.

What is not reported about the statistical analysis?

The record identifies nine outcome measures and lists one analysis, but does not give enough detail to reconstruct the analysis strategy. The statistical methods, any adjustment for baseline characteristics, the handling of repeated assessments, and the planned approach to confidence intervals are not reported in the registry record. Nor are the analysis software, model diagnostics, or rules for protocol deviations reported. These omissions do not show that such procedures were absent from the full study conduct; they mean they cannot be described from the registry record.

Over-dispersion is a general concern when count data vary more than a chosen statistical model expects. It can lead to uncertainty estimates that are too narrow if not accommodated. The primary outcome is described as a participant count rather than a rate with reported follow-up exposure, and no model or variance estimate is provided. There is not enough information to assess whether over-dispersion was relevant to any analysis or how it was handled. Statistical interpretation should therefore remain tied to the reported design and outcomes, without inventing model details.

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("Methoxsalen With ECP")
n      <- c(29)
events <- c(16)
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  Methoxsalen With ECP  29  16  55.2  35.7  73.6  37.5  71.6  37.4  71.7

11. Multiplicity, Missing Data and Analysis Populations

The record lists nine outcome measures, including the primary Week 4 response endpoint and secondary outcomes assessed at Week 8 and Week 12, measures of response duration, other response criteria, steroid dose, organ-stage assessments, and adverse events. When a study examines several outcomes or assessment times, the chance of at least one apparently unusual result can increase if each is tested separately. In general, multiplicity procedures may control that risk, prioritize selected outcomes, or distinguish confirmatory from exploratory analyses. The statistical plan and any multiplicity adjustment are not reported in the registry record.

The registry record lists one analysis but does not explain which outcome or analysis it refers to, whether the listed outcomes were tested, or whether any testing hierarchy was prespecified. It does not report adjusted or unadjusted p-values for the primary endpoint. No conclusion about multiplicity control can be drawn from the number of listed outcomes alone.

Missing-data handling is also not reported. The record does not state how participants who stopped treatment, missed an assessment, or lacked evaluable organ information were classified for the Week 4 response outcome. It does not give a missingness count, imputation method, or sensitivity analysis. Because the primary population is described as all participants, details of the analysis denominator and inclusion of unavailable outcomes would be needed to understand how that label was implemented.

The primary population is identified as all participants, but other analysis populations—such as safety or per-protocol populations—and their definitions are not reported in the registry record. The enrollment entry should not be taken to establish that all participants contributed data to every outcome. The available information therefore does not allow a full assessment of how incomplete follow-up, multiple endpoints, or analysis-population choices may affect the reported findings.

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: pyrexia (6.9%), acute graft versus host disease (6.9%), aspergilloma (3.4%), bk virus infection (3.4%), cellulitis (3.4%).

For Methoxsalen With ECP, the registry record reports 12 participants with serious adverse events among 29 at risk, 17 participants with other adverse events among 29 at risk, and 3 deaths. These are separate descriptive safety figures. The record does not provide a causal attribution for the events or deaths, so they should not be interpreted as counts established to have resulted from the study regimen.

The most frequently listed serious adverse-event terms are pyrexia and acute graft versus host disease, each reported as 6.9%. Aspergilloma, Bk virus infection, cellulitis, and cystitis viral are each reported as 3.4%. These percentages describe the registry’s listed terms; the record does not provide additional event-level detail such as timing, duration, recurrence, or severity grading beyond the serious-event classification. A term-level percentage should not be read as a treatment-attributable risk.

In general, adverse-event categories can overlap: a participant may experience more than one event, and a term count is not necessarily the same as a participant count. The registry record does not give enough detail to determine how the listed terms relate to the participant counts. The deaths are reported separately, but their timing and relationship to adverse events or treatment are not reported in the registry record. These figures are safety observations, not a comparative estimate against an untreated or alternative-treatment group.

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

The available information establishes that the study evaluated methoxsalen with ECP in pediatric participants with steroid-refractory aGvHD and defined a primary Week 4 overall-response outcome using the modified IBMTR Severity Index. It also records selected adverse-event counts and the study’s terminated status. These registry entries describe the design and the material reported in the registry.

The primary response count itself is not reported in the registry record. Consequently, the results do not quantify how many participants met the Week 4 response definition, how precise any response estimate might be, or whether outcomes differed from a comparator. No efficacy conclusion can be drawn from an endpoint definition alone. The secondary response and disease-assessment outcomes are listed, but their results are not reported in the registry record.

Even a reported response count in this single-arm study would not establish that the regimen caused the response. Without a concurrent comparator, the outcome in treated participants cannot be contrasted directly with what would have occurred under another approach. The open-label design also means masking was not used, while the record does not describe independent or blinded assessment. These limitations concern the strength of causal interpretation; they do not prove that the assessments were inaccurate.

The serious and other adverse-event counts, listed serious-event terms, and deaths describe safety observations in the study group. They do not determine whether the regimen caused any event, and the registry record does not give event-level timing or causality assessments. The terminated status likewise does not explain why the study ended. The reason for termination and its implications for completion are not reported in the registry record.

Sample-size illustrations based on an observed effect are planning tools under their stated assumptions, not evidence of observed power or a substitute for missing trial results. The stated assumptions describe a hypothetical two-arm design, unlike this single-arm trial. The registry record does not report the trial’s original sample-size justification or the primary outcome result needed to interpret an effect-based planning illustration. Readers should therefore distinguish the recorded design and safety counts from unreported efficacy conclusions.

14. Why This Trial Matters Statistically

This study raises a central statistical distinction between describing outcomes and estimating a treatment effect. A single-arm study can document whether participants receiving a regimen meet an outcome definition, and it can characterize safety observations within the group. A treatment effect, however, is a contrast between outcomes under different treatment conditions. Without a concurrent comparator, that contrast is not directly observed. This distinction is especially important when disease status and care may change over time.

The endpoint definition also illustrates why outcome construction matters. Overall response combines complete and partial response, while requiring that next-line systemic treatment not be added under the stated criteria. The classification relies on organ-level signs, symptoms, and stage changes. Interpretation therefore depends not only on the response count but also on consistent assessments, evaluability, and transparent handling of unavailable data. Those analysis details are not reported in the registry record.

Several assessment times and outcome types are listed, creating a need to understand which analyses were intended to be confirmatory and how multiplicity was handled. The statistical plan is not reported, so the status of secondary comparisons cannot be established. Similarly, the record’s terminated status is relevant context but does not reveal why the trial ended or whether the planned analysis was completed.

Finally, the registry record illustrates the difference between an outcome being specified and its result being reported. Here, the primary response endpoint is defined, but its observed count is not reported. That gap prevents readers from judging the size and precision of the primary finding. The safety figures remain interpretable as descriptive registry entries, with important limits on attribution and comparison.

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 primary limitation for interpreting efficacy is that the Week 4 response count is not reported in the registry record. The record also does not report a primary response proportion, confidence interval, p-value, or treatment-effect estimate. This prevents assessment of the magnitude and precision of the primary outcome, regardless of the endpoint’s detailed definition.

The design is single-arm and open-label, with no concurrent control. Outcomes therefore cannot be compared directly with those under another intervention, and causal attribution is limited. The registry record does not describe whether response assessors were independent or blinded, or how organ-level assessments were standardized. These details are particularly relevant to an endpoint based on clinical classification across evaluable organs.

Participant flow and completeness of assessment are not reported. The record does not state how many participants completed treatment or follow-up, how many had evaluable Week 4 outcomes, or how missing assessments were handled. The analysis populations beyond the primary population label are not reported. These omissions limit understanding of the denominator and the potential influence of unavailable outcome data.

The study is recorded as terminated, but the reason for termination and its effect on completion of planned assessments are not reported in the registry record. The statistical analysis plan, sample-size justification, multiplicity approach, and details of the listed analysis are also not reported. The provided safety figures lack event timing and causal attribution, and they do not offer a comparator-based safety contrast. These limitations leave the record useful for describing the intended design and selected safety observations, but insufficient for a full evaluation of efficacy, precision, or causality.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Methoxsalen and Extracorporeal Photopheresis (ECP) for the Treatment of Pediatric Participants With Steroid Refractory Acute Graft Versus Host Disease — NCT02524847.
• Registry record last updated 2020-09-11; status last verified 2020-07.
• Results first posted 2020-09-11.
• 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

NCT02524847 is a Phase 3, multicenter, open-label, single-arm study of methoxsalen with ECP in pediatric participants with steroid-refractory acute graft-versus-host disease. Enrollment is recorded as 29, and status is terminated. The primary endpoint is the number of participants achieving overall response by the modified IBMTR Severity Index at Week 4; the response count is not reported in the registry record.

The regimen is listed as methoxsalen 20 µg/ml with ECP, three times per week for Weeks 1 to 4 and two times per week for Weeks 5 to 12. The record reports 12 participants with serious adverse events among 29 at risk, 17 with other adverse events among 29 at risk, and 3 deaths. It does not report the reason for termination, primary efficacy results, or the detailed statistical and missing-data methods needed for fuller interpretation.