This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. Numerical trial results on this page are restricted to the results posted in the ClinicalTrials.gov record; calculations added by Clinical Biostats are labelled as such.
1. Trial at a Glance
This page covers the registry results for Infliximab and Intravenous Immunoglobulin Therapy in Treating Patients With Steroid-Refractory Pneumonitis (NCT04438382), a randomized phase 2 study. The primary outcome is Pneumonitis Response Rate.
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT04438382 |
| Phase | Phase 2 |
| Condition | Hematopoietic and Lymphoid Cell Neoplasm, Malignant Solid Neoplasm, Steroid-Refractory Pneumonitis |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | None (open label) |
| Enrollment | 1 |
| Arms in the primary outcome | n/a |
| Lead sponsor | ECOG-ACRIN Cancer Research Group |
| Results posted | Yes |
2. Clinical Question
Population
People with steroid-refractory pneumonitis in the setting of anti-PD-1/PD-L1 treatment were the intended population. The registry also lists hematopoietic and lymphoid cell neoplasms and malignant solid neoplasms among the conditions. Only one participant was enrolled, and individual data were not reported in the registry record.
Intervention
Arm A used intravenous infliximab on day 1, followed by a prednisone taper given intravenously or orally for 4-6 weeks, absent disease progression or unacceptable toxicity. A further infliximab dose on day 14 could be given at the treating physician’s discretion.
Comparator
Arm B used intravenous immunoglobulin therapy over 2-5 days according to institutional guidelines, followed by a prednisone taper given intravenously or orally for 4-6 weeks, absent disease progression or unacceptable toxicity. Both arms therefore included a prednisone taper; the registry record does not report which arm received the enrolled participant.
Primary question
The primary question concerned pneumonitis response at day 28. Response was defined by an improvement of >= 20% in the PaO2/FiO2 measure, comparing day 28 with day 1, using PaO2 and the FiO2 received at the time of the arterial blood gas assessment. The registry record does not report a response result or a comparative treatment estimate.
This question concerns a specific endpoint in a specific clinical context: response of pneumonitis that did not respond to steroid treatment. It is not a general comparison of the medicines across other diseases or settings. The intended randomized, parallel-arm structure would ordinarily support comparison between the two assigned strategies, but the record does not provide an arm-level outcome result. With only one participant enrolled and individual data withheld for confidentiality, the posted information cannot show whether the endpoint was met or whether the arms differed. The clinical question is recorded, but a treatment comparison cannot be estimated from the published registry results.
3. Trial Design
NCT04438382, titled “Infliximab and Intravenous Immunoglobulin Therapy in Treating Patients With Steroid-Refractory Pneumonitis,” is listed as a phase II randomized, parallel-arm study. Masking is listed as none, so the trial was open-label. The sponsor was ECOG-ACRIN Cancer Research Group. The registry lists a start date of 2021-01-07 and a primary completion date of 2023-12-21. Its current status is terminated.
The two strategies differed in the immunosuppressive treatment preceding a prednisone taper. Arm A received infliximab intravenously on day 1; an additional dose on day 14 was permitted at the treating physician’s discretion. Arm B received intravenous immunoglobulin over 2-5 days according to institutional guidelines. Both arm descriptions specify a prednisone taper, given intravenously or orally for 4-6 weeks, unless disease progression or unacceptable toxicity intervened. The registry lists prednisone and methylprednisolone among the interventions, while the arm descriptions specify prednisone.
Randomization assigns participants to study arms according to a chance-based process. In general, this can reduce systematic differences between groups, though it does not guarantee that groups will be comparable, particularly when enrollment is very small. An open-label design means treatment assignment is not masked. That can matter for outcomes influenced by decisions or assessments, although the registry record does not report how endpoint assessment was conducted or whether assessors were independent.
Enrollment was one participant. The registry record does not report the allocation sequence, the participant’s assigned arm, or whether any treatment was received. It also does not report a planned sample-size justification, the reason for termination, or any interim-analysis plan. The registry lists ten outcome measures and names secondary outcomes concerning radiologic response, spirometry, diffusion capacity, oxygen saturation, deaths within 28 days, adverse events, and the Functional Assessment of Cancer Therapy - Lung Version 4 total score. Those outcome listings describe planned assessments, not evidence that usable results were recorded.
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
| Stage | Arm A (Infliximab) | Arm B (Intravenous Immunoglobulin Therapy) |
|---|---|---|
| STARTED | 0 | 1 |
| COMPLETED | 0 | 1 |
| NOT COMPLETED | 0 | 0 |
Analysis population for the primary outcome, as defined in the registry: Only one patient was enrolled on this study. Due to confidentiality reasons, individual data was not reported.
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: Only one patient was enrolled on this study. Due to the concern on confidentiality, individual data is not reported due to confidentiality reasons.
| Characteristic | Unit | Arm A (Infliximab) | Arm B (Intravenous Immunoglobulin Therapy) | Total |
|---|---|---|---|---|
| Participants analysed | 0 | 0 | 0 | |
| Age, Categorical: <=18 years | — | — | — | |
| Age, Categorical: Between 18 and 65 years | — | — | — | |
| Age, Categorical: >=65 years | — | — | — | |
| Sex: Female, Male: Female | — | — | — | |
| Sex: Female, Male: Male | — | — | — | |
| Ethnicity (NIH/OMB): Hispanic or Latino | — | — | — | |
| Ethnicity (NIH/OMB): Not Hispanic or Latino | — | — | — | |
| Ethnicity (NIH/OMB): Unknown or Not Reported | — | — | — | |
| Race (NIH/OMB): American Indian or Alaska Native | — | — | — | |
| Race (NIH/OMB): Asian | — | — | — | |
| Race (NIH/OMB): Native Hawaiian or Other Pacific Islander | — | — | — | |
| Race (NIH/OMB): Black or African American | — | — | — | |
| Race (NIH/OMB): White | — | — | — | |
| Race (NIH/OMB): More than one race | — | — | — | |
| Race (NIH/OMB): Unknown or Not Reported | — | — | — |
6. Primary Endpoint
The registry lists the primary outcome as Pneumonitis Response Rate, measured over At day 28.
| Element | Registry entry |
|---|---|
| Result type | N/A |
| Measure of spread | not applicable |
| Unit | not stated |
| Endpoint classification | not reduced to a single comparison |
The primary endpoint was pneumonitis response rate at day 28. The registry defines response as an improvement of >= 20% in the PaO2/FiO2 measure on day 28 compared with day 1. PaO2 is the partial pressure of arterial oxygen, and FiO2 is the fraction of inspired oxygen. The definition uses the PaO2 measurement and the FiO2 the participant received when the arterial blood gas assessment was made.
This endpoint combines an oxygenation measure with a specified comparison between two assessment days. In general, a response threshold turns a continuous measurement into a categorical outcome: participants either meet the stated definition or do not. That can make a result easier to interpret as a response proportion, but it also means that observations near the cutoff can fall on opposite sides of the category despite being similar in magnitude. The registry record does not report additional details on endpoint adjudication, handling of unavailable measurements, or the precise computational convention used for the percentage improvement.
The outcome was intended to assess early change in oxygenation after treatment. A response proportion would ordinarily require the number meeting the definition and the number assessed in each relevant group. Neither an arm-specific response count nor an overall response count is reported. The registry states that only one participant was enrolled and that individual data were withheld for confidentiality. Consequently, the endpoint definition is available, but an observed response rate or between-arm comparison is not.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The central statistical fact is that enrollment was one participant. Although the study was randomized and designed with parallel arms, the registry record does not report the assigned arm or an arm-specific result. It also states that individual data were not reported for confidentiality. The primary outcome therefore has no posted response count, response rate, or comparative estimate. No conclusion about a difference between infliximab and intravenous immunoglobulin therapy can be drawn from the registry results.
The record lists zero participants at risk and zero affected for the serious and other adverse-event categories in both arms, and zero deaths. Those entries should not be read as evidence that no adverse events or deaths occurred among treated participants. The recorded at-risk counts are also zero, and the record reports only one enrollee overall. Without an interpretable exposure denominator or participant-level information, the entries do not establish the safety experience of either strategy.
The registry identifies ten outcome measures and names secondary outcomes involving radiologic response, pulmonary function, oxygen saturation, deaths within 28 days, adverse events, and a lung-related functional assessment score. However, the analysis count is zero. The record does not provide estimates, uncertainty intervals, p-values, or participant-level outcomes for these measures. A listed endpoint is not itself a reported result.
In general, a confidence interval describes the range of values compatible with an estimate under a statistical model and its assumptions. No confidence interval is reported for the primary endpoint. With no posted response estimate and no comparative analysis, there is no interval to interpret. The same limitation applies to significance testing: a p-value would require an analyzed comparison, which the registry record does not report.
The record gives dates, design features, intervention descriptions, and an enrollment count, but not a reason for termination or an explanation of how enrollment ended at one. It does not report whether the participant completed follow-up, whether endpoint measurements were available, or whether the study stopped before planned recruitment. These gaps prevent distinguishing among possible explanations for the absence of results. The statistical interpretation must therefore remain descriptive: a randomized phase II comparison was registered, one participant was enrolled, and no analyzable primary or comparative outcome result is posted.
9. Statistical Methods Explained
Methods are described only to the extent that they appear in the registry record. It identifies a randomized, parallel-arm, open-label phase II study and defines a primary response endpoint. It does not report a detailed statistical analysis plan. The general explanations below describe what these design and analysis concepts mean; they do not imply that a particular unreported method was used.
What does randomization contribute?
Randomization assigns participants to treatment groups using a chance-based process. In general, its purpose is to limit systematic allocation bias and to support a fair comparison of groups. Randomization does not guarantee balance in every characteristic, and its ability to produce comparable groups depends in practice on the amount of enrollment. With one participant enrolled, the record does not support an informative comparison between the two assigned strategies. The participant’s allocation is not reported, so it cannot be determined which treatment strategy was represented in the available enrollment.
A parallel-arm design means participants are assigned to one study strategy rather than planned to receive both strategies in sequence. The registry lists Arm A as infliximab followed by a prednisone taper, with a discretionary additional infliximab dose, and Arm B as intravenous immunoglobulin followed by a prednisone taper. The registry does not describe a crossover procedure. The record also does not report the randomization method, allocation concealment, or any stratification factors; these details are not reported in the registry record.
What does open-label mean for interpretation?
Masking is listed as none. In general, an open-label study is one in which treatment assignment is known rather than concealed from participants or study personnel. Knowledge of assignment can influence care decisions, reporting, or outcome assessment, depending on the endpoint and study procedures. The primary endpoint uses oxygenation measurements, but the registry record does not report whether measurements were centrally reviewed, whether endpoint assessors were masked, or how assessment procedures were standardized. It would therefore be inappropriate to assume either that assessment bias occurred or that safeguards eliminated it.
How should the primary response measure be understood?
The primary endpoint is a binary response classification based on a threshold: an improvement of >= 20% in PaO2/FiO2 at day 28 compared with day 1, with PaO2 and the FiO2 received at the time of arterial blood gas assessment used for the measure. In general, a threshold-based endpoint can be summarized as the proportion of participants who meet its definition. It also depends on the availability and comparability of measurements at the specified assessment times. The registry record does not describe rules for missing measurements, repeat tests, or values obtained outside the planned assessment times.
The posted record does not report the number who met the threshold, the number with an endpoint assessment, or the result by arm. The outcome description also does not report a statistical model or an estimation method for the response rate. No effect estimate, standard error, confidence interval, or hypothesis-test result appears in the registry record.
What can be said about sample-size calculations?
The registry record does not report the trial’s own sample-size justification, target enrollment, or assumptions used when designing the study. It also does not report an observed treatment effect from which a meaningful planning illustration could be interpreted. General planning illustrations based on an observed effect are not the trial’s design rationale and are not evidence of study power. A sample-size calculation is a planning exercise based on chosen assumptions; it cannot replace an observed treatment estimate. “Observed power” is not a useful substitute for reporting an effect estimate and its uncertainty.
For a binary endpoint, planning generally depends on assumptions about response rates and the difference the study is intended to detect, together with the chosen error rates and allocation. Those inputs are not reported as the trial’s design assumptions. No sample-size justification should be inferred from the one participant enrolled. That count describes enrollment, not the number required to answer the research question reliably.
Why does over-dispersion matter in general?
Over-dispersion describes variation greater than a statistical model expects. It can matter for outcomes such as counts or rates because a model that understates variation may produce uncertainty estimates that are too narrow. The registry’s named outcomes include event counts and functional measures, but no fitted model, dispersion estimate, or analysis result is reported. There is therefore no basis to claim that over-dispersion was assessed or to explain its effect on this trial’s estimates. This is a general consideration, not a finding about the study.
What does the analysis record show?
The registry lists zero analyses. It does not report an analysis population, such as all randomized participants or participants who received treatment, and it does not describe exclusions from analysis. Nor does it report an approach to missing outcome data, multiplicity, covariate adjustment, or interim monitoring. These are important decisions in a clinical-trial analysis, but their details are not reported in the registry record. Without a posted analysis and outcome data, no statistical model or inferential result can be reconstructed from the registered endpoint description alone.
10. Multiplicity, Missing Data and Analysis Populations
The registry names multiple outcomes: the primary pneumonitis response rate and secondary assessments involving radiologic response, spirometry, diffusion capacity, oxygen saturation, deaths within 28 days, adverse events, and the Functional Assessment of Cancer Therapy - Lung Version 4 total score. It also reports ten outcome measures overall. When a study assesses multiple outcomes, multiplicity can arise from testing several hypotheses or examining several measures. In general, multiple testing can increase the chance of at least one apparently unusual result if every test is interpreted without accounting for the number of tests. A prespecified strategy may identify one primary outcome or adjust interpretation across outcomes. The registry record does not report a multiplicity strategy or a hierarchy for interpreting secondary outcomes.
The record reports that individual data were not reported because only one participant was enrolled and confidentiality was a concern. It does not report whether day 1 or day 28 measurements were collected, whether any outcome was missing, or how missing values would have been handled. The absence of an outcome result should not be treated as a measured value, a failure to respond, or a documented loss to follow-up. Missing-data procedures are not reported in the registry record.
The analysis population is also not specified. The registry does not say whether analyses were planned for all randomized participants, those who received treatment, or another defined group. It does not report the participant’s assigned arm or treatment exposure. The analysis count is zero, and no participant-level outcomes or arm-level estimates are posted. Thus, neither the handling of missing data nor the population included in any efficacy or safety analysis can be established from the record.
11. What the Result Does, and Does Not, Mean
The record establishes that a randomized, open-label phase II study was registered to compare two treatment strategies for steroid-refractory pneumonitis. It describes the arms, identifies pneumonitis response at day 28 as the primary endpoint, and reports enrollment of one participant. It also states that individual data were not reported for confidentiality. These registry entries describe the study’s registration and the extent of the posted results; they do not establish an outcome for either strategy.
The posted information does not show whether the participant met the primary response definition, which arm the participant entered, or whether the participant completed the planned assessments. It does not provide a response rate, comparative effect, uncertainty interval, or statistical test. The absence of those results means the registry record cannot answer whether the outcomes differed between arms. It also cannot establish that either strategy was effective or ineffective.
The adverse-event entries require similar restraint. With zero participants recorded at risk in both arms, the listed zeros cannot be interpreted as a safety rate or as evidence that no adverse events occurred. There is no reported treatment-exposure denominator or arm assignment with which to interpret the entries. They do not support a comparative safety conclusion.
In general, randomized comparisons can help estimate treatment differences when there are adequate outcome data and a defined analysis. Randomization alone does not generate an estimate, and a registered outcome definition is not a result. In this record, the gap between the planned question and the posted evidence is especially consequential because enrollment was one participant and individual outcomes were withheld. Any broader conclusion about the intervention strategies would go beyond what the registry record reports.
12. Why This Trial Matters Statistically
This study illustrates the difference between a registered clinical question and a statistically answerable comparison. The record describes randomization, two treatment strategies, and a defined primary endpoint, but the enrollment count is one and no outcome analysis is reported. A carefully specified endpoint cannot by itself produce an estimate; the comparison also requires reported outcome data and an analysis that accounts for the study design.
It also highlights how design labels should be interpreted in context. Randomized and parallel-arm describe the intended comparison structure, while open-label describes masking. Those features do not guarantee adequate information, and they do not tell readers which arm the enrolled participant received. With the assignment and outcome data unavailable, the design cannot be used to infer a treatment contrast.
The record’s confidentiality statement marks a further statistical tension: protecting individual information can limit what can be shared, while estimates and uncertainty are needed to evaluate results. Here, the registry does not provide an aggregate primary outcome or comparative result alongside the enrollment information. That leaves readers unable to assess the magnitude or precision of an effect.
Finally, the entries for adverse events demonstrate why denominators matter. A count recorded as zero is not interpretable without knowing who was at risk and how exposure was captured. The record lists zero participants at risk, so the accompanying zero event counts do not convey comparative safety experience. These are general lessons about evidence interpretation, not conclusions about the treatments themselves.
13. 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 main limitation is the very small enrollment: one participant. The registry does not report the assigned arm, an arm-level outcome, or a comparative estimate. It states that individual data were withheld for confidentiality. As a result, the primary response endpoint cannot be summarized from the posted information, and no treatment contrast or uncertainty estimate is available.
The study status is terminated, but the registry record does not report why it was terminated. It also does not report the planned enrollment target, the study’s sample-size justification, or whether termination occurred before planned recruitment was complete. These omissions prevent readers from determining the intended scale of the study or the reason the record contains so little outcome information.
For the primary endpoint, the registry provides a threshold definition and assessment days, but it does not report assessment results, endpoint adjudication procedures, or rules for missing measurements. It does not report whether the participant received treatment or completed follow-up. The allocation sequence and arm assignment are not reported, and the analysis population is unspecified.
The registry lists adverse-event categories, but both arms show zero participants at risk. The corresponding zero counts therefore do not establish an absence of events or allow safety rates to be compared. The record also does not describe how adverse events were collected or analyzed. More generally, the analysis count is zero, and the statistical plan, multiplicity handling, interim analyses, and missing-data methods are not reported in the registry record.
14. Related Tutorials
15. Related Statistical Calculators
16. Sources
- ClinicalTrials.gov: Infliximab and Intravenous Immunoglobulin Therapy in Treating Patients With Steroid-Refractory Pneumonitis — NCT04438382.
- Registry record last updated 2026-09-29; status last verified 2026-09.
- Results first posted 2024-03-26.
- 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.
Continue through the Clinical Biostats statistical pathway
Use this trial's endpoint as a starting point for deeper study of effect measures, confidence intervals, and sample-size planning.
17. Record Summary
NCT04438382 is a randomized, open-label, parallel-arm phase II study sponsored by ECOG-ACRIN Cancer Research Group. It compared an infliximab strategy with intravenous immunoglobulin therapy for steroid-refractory pneumonitis; both arms included a prednisone taper. The registry lists the start date as 2021-01-07, the primary completion date as 2023-12-21, and the status as terminated.
The primary endpoint was pneumonitis response at day 28, defined as an improvement of >= 20% in PaO2/FiO2 compared with day 1. The registry reports one participant enrolled and states that individual data were not reported for confidentiality. No primary outcome result, arm-level comparison, or analysis is reported. The listed adverse-event entries show zero participants at risk in each arm, so they do not establish a safety rate or absence of events. The reason for termination and the statistical analysis plan are not reported in the registry record.