Lipopolysaccharide (LPS) or Granulocyte-Macrophage Colony-Stimulating Factor (GM-CSF) Challenge Study on Healthy Subjects: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Lipopolysaccharide (LPS) or Granulocyte-Macrophage Colony-Stimulating Factor (GM-CSF) Challenge Study on Healthy Subjects (NCT03306589), a randomized phase 1 study. The primary outcome is Part 1: Change From Baseline Primary Soluble Inflammatory Mediators in Blood: Tumor Necrosis Factor (TNF) Alpha and Interleukin (IL) 6 for LPS Arm.

2. Clinical Question

3. Trial Design

NCT03306589 is recorded as a completed, randomized, open-label, parallel-group Phase 1 study sponsored by GlaxoSmithKline. The registry lists enrollment of 12. The recorded start date is 2017-08-11, and the primary completion date is 2018-01-03. The study description presents a two-part design: Part 1 for dose exploration and Part 2 as a continuation intended to obtain more precise outcomes. The record does not describe the statistical basis for that continuation.

Participants were assigned to receive LPS or GM-CSF. The LPS arm description gives dose escalation ranging from 0.5 nanogram (ng)/kg to 4 ng/kg; the GM-CSF arm description gives a range from 5 to 15 microgram (µg)/kg. The description says each part included two sessions. In the first session, participants received a challenge agent and had cantharidin-induced blisters, followed by blood draws and blister collection at stated post-induction times. Participants returned after a minimum 14-day blister-healing period for the second session. Part 1 describes up to 6 cohorts and dose escalation until a well-tolerated dose was determined.

The registry labels the study as unmasked. It does not report how the randomization sequence was generated, how allocation was concealed, or whether any safeguards were used to reduce measurement bias in this open-label setting. It also does not report the planned sample-size justification or a detailed statistical analysis plan. The brief description ends mid-sentence when discussing how the same dose would be used, so the remaining continuation details are not reported in the registry record.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Before assignment: Part 2 of the study was not conducted as agreed criteria for Interim analysis was achieved.

Part 1 (69 Days)

Part 2 (69 Days)

Analysis population for the primary outcome, as defined in the registry: Safety Population. Only those participants with data available at the specified data points were analyzed (represented by n= X in the category titles).

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: Part 2 of the study was not conducted as agreed criteria for Interim analysis was achieved.

Only the first 6 of 7 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as Part 1: Change From Baseline Primary Soluble Inflammatory Mediators in Blood: Tumor Necrosis Factor (TNF) Alpha and Interleukin (IL) 6 for LPS Arm, measured over Baseline, Session 2: -5, 10, 25, 40 minutes, 1 hour 10 minutes, 1 hour 40 minutes, 2 hours 40 minutes,5 hours 40 minutes on Day 1. Pre-fluid sample on Day 2 and Day 3.

The listed primary outcome is change from baseline in the soluble inflammatory mediators TNF-alpha and IL-6 in blood for the LPS arm in Part 1. These are measured in picograms per milliliter. Blood samples were scheduled at baseline and at multiple timepoints during Session 2, including measurements from minutes after challenge through later assessments and pre-fluid samples on Day 2 and Day 3. The registry gives the timepoints as Baseline, Session 2: -5, 10, 25, 40 minutes, 1 hour 10 minutes, 1 hour 40 minutes, 2 hours 40 minutes, 5 hours 40 minutes on Day 1, and pre-fluid sample on Day 2 and Day 3.

In general, measuring change from baseline describes how an individual's post-challenge value differs from the value used as their starting reference. The registry defines baseline as the latest pre-challenge LPS assessment with a non-missing value, including assessments from unscheduled visits; change is the specified visit value minus that baseline. The mean and standard deviation are the stated summary measures. Biomarkers are used in challenge studies to characterize biological responses over time. The registry describes measurement but does not establish a threshold for interpreting a change as clinically meaningful.

The endpoint is limited to blood TNF-alpha and IL-6 in the LPS arm for Part 1. Other listed outcomes include blister biomarkers, blister volume, cell numbers, and cell activation markers. The numerical estimates for the primary endpoint are not reported in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry identifies the primary endpoint as a continuous measure summarized by a mean and standard deviation. A continuous outcome can take measured values along a scale, here reported in picograms per milliliter. The mean describes the central level of the observed values, while the standard deviation describes their spread around that mean. These summaries answer different questions: a mean alone does not show how variable observations were, and a standard deviation alone does not give the direction or size of an average change.

The primary measure is change from baseline. The record defines baseline as the latest pre-challenge LPS assessment with a non-missing result, including results from unscheduled visits. At each specified visit, the baseline value is subtracted from the value at that visit. This approach makes each participant's pre-challenge measurement the reference for describing subsequent change. It does not remove all possible sources of variation, and comparisons of change still depend on the observations available at the relevant times and on the appropriateness of the chosen baseline.

Measurements were scheduled at multiple points in Session 2, beginning with baseline and including several assessments during Day 1, as well as pre-fluid samples on Day 2 and Day 3. Repeated measurements can describe the time course of a response, but they also create multiple results to interpret. In general, measurements from the same participant over time are related rather than independent. A model that uses the repeated structure can account for that correlation; the specific model used, if any, is not reported in the registry record. The record does not describe a method for estimating or testing changes across the full time course.

The endpoint description states that only participants with data available at the specified data points were analyzed, with the available count represented in category titles. It also notes that NA indicates data were unavailable for calculating a standard deviation for a single participant. These statements indicate that the number contributing to an analysis may vary by measurement. The numerical category counts and the reasons for missing observations are not reported in the registry record. In general, a summary based only on available observations can be affected if the participants with missing data differ in relevant ways from those with observed data. The record does not report a missing-data model, imputation method, or sensitivity analysis.

The description says that all participants randomized to receive treatment and receiving one dose of challenge agent were included in the Safety Population. This defines a treated safety group, but it is not a complete description of the analysis population for every efficacy outcome. The primary endpoint text separately restricts analyses to participants with data at the specified data points. The registry does not report whether a different population or rule was used for other outcomes.

Multiplicity is relevant because the study lists 36 outcome measures and includes several biomarkers, timepoints, and assessment domains. In general, evaluating many outcomes or timepoints increases the chance of observing a seemingly notable result by chance if each is tested separately. A multiplicity adjustment can control that issue for a defined family of tests, but no adjustment method or family of primary comparisons is reported in the registry record. The record also reports 0 analyses, and does not provide inferential results for the primary outcome. A planned statistical method, including whether estimates were intended to be adjusted, is not reported in the registry record.

Randomization can reduce systematic differences between assigned groups on average, but it does not guarantee balance in a small study. Open-label conduct means participants and study personnel were not masked according to the registry. For laboratory biomarkers, objective measurement may limit some forms of subjective assessment, but the record does not describe laboratory blinding, assay procedures, or other protections against measurement bias. Randomization procedures and allocation concealment details are also not reported.

The registry's sample-size assumptions describe a planning illustration based on the observed effect: a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. The effect to detect is described as a stated fraction of the observed effect. For time-to-event tables, the stated output is total events; for other outcomes, it is participants per arm. These are planning assumptions, not observed power or evidence that the trial was designed around that calculation. The primary outcome is not reduced to one comparison, and the observed effect and planning values are not reported in the registry information. The study's own sample-size justification is not reported in the registry record.

In general, confidence intervals help show the range of values compatible with an estimate under a statistical model, while a p-value summarizes how unusual data would be under a specified null hypothesis and assumptions. Neither should substitute for effect size, uncertainty, design context, or clinical relevance. The registry does not report primary-endpoint confidence intervals or p-values, nor does it report an inferential model. Accordingly, the method that generated any primary outcome estimate cannot be evaluated from the registry record.

10. Multiplicity, Missing Data and Analysis Populations

The study lists 36 outcome measures across inflammatory biomarkers, blister volume, cell numbers, and cell activation markers. The primary endpoint itself covers TNF-alpha and IL-6 at multiple scheduled timepoints. In general, testing many outcomes or timepoints creates a multiplicity problem: the more separate comparisons examined, the more opportunities there are for chance findings. A prespecified hierarchy, a defined primary comparison, or an adjustment procedure can help manage interpretation. The registry record does not report such a hierarchy or a multiplicity adjustment method.

The primary outcome is not reduced to one comparison in the registry information. Numerical outcome estimates and inferential results are not reported in the registry record, and the record lists 0 analyses. It is therefore not possible to determine how many comparisons were actually tested, whether any outcomes were selected after examining the data, or whether any adjustment was applied. Those details are not reported in the registry record.

Missingness is also relevant. The endpoint description says that only participants with data at each specified data point were analyzed, so the number contributing may vary across timepoints. The record notes that an unavailable value could prevent a standard deviation from being calculated for a single participant. It does not report the pattern or reasons for missing biomarker measurements, the number analyzed at each timepoint, or a strategy for handling missing data beyond analyzing available observations. In general, available-case summaries can be difficult to interpret if missingness is related to participant characteristics or outcomes.

The Safety Population includes participants randomized and given a dose of challenge agent. The primary biomarker analysis, by contrast, is described as including only those with data available at the specified points. Whether other outcomes used the same population definition is not reported in the registry record. No sensitivity analyses or missing-data assumptions are described.

11. 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 0% reporting threshold.

Most frequent non-serious adverse event terms

The serious adverse event figures are reported separately by listed group. For Part 1: LPS 0.5 ng/kg, 2 were at risk and 0 had serious adverse events; for Part 1: LPS 0.75 ng/kg, 4 were at risk and 0 had serious adverse events; for Part 1: LPS 1 ng/kg, 2 were at risk and 0 had serious adverse events; and for Part 1: GM-CSF 60 µg/m^2, 4 were at risk and 0 had serious adverse events. Deaths are recorded as 0 for each of those groups. No serious-event terms are listed.

The same registry record lists Part 2 groups for LPS and GM-CSF with 0 at risk, 0 with serious adverse events, and 0 deaths. Because no participants are recorded at risk in those Part 2 categories, those zeros should not be read as evidence that exposed participants in Part 2 had no serious events.

Other adverse events are distinct from serious adverse events. The record reports 2 affected among 2 at risk in the LPS 0.5 ng/kg group, 4 among 4 in the LPS 0.75 ng/kg group, 2 among 2 in the LPS 1 ng/kg group, and 3 among 4 in the GM-CSF 60 µg/m^2 group. No event-term breakdown is included. These descriptive counts do not establish a comparative safety profile. The GM-CSF dose category is stated in µg/m^2 here, while its arm description gives a dose range in µg/kg; the registry record does not explain this difference in units.

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

This study is an exploratory investigation of methods for exposing healthy participants to LPS or GM-CSF as systemic challenge agents and measuring inflammatory responses. Its outcomes include blood biomarkers, blister assessments, and cellular measures. The primary outcome named in the record focuses on changes in blood TNF-alpha and IL-6 for the LPS arm in Part 1. This is a biological measurement question, not a direct assessment of symptoms, function, or clinical benefit.

The record establishes that the study was randomized, open-label, and parallel-group, and describes dose escalation, blister induction, sampling schedules, and safety reporting. It does not report numerical primary outcome estimates. As a result, the registry information does not show how large or variable the measured biomarker changes were, whether patterns differed across timepoints, or how precisely any effect was estimated. The intervention descriptions and the existence of scheduled sampling are not evidence on their own that a particular response occurred.

The study also does not provide a reported direct statistical comparison of LPS with GM-CSF on the primary endpoint. The primary endpoint is specifically for the LPS arm, and the record does not state a formal comparative hypothesis or decision rule. The existence of randomized groups should not be taken to mean that every listed outcome was analyzed as a head-to-head comparison.

Part 1 safety counts are descriptive and come from small dose groups. The Part 2 categories record no participants at risk, so their zero event counts cannot establish that the agents were free of serious events in an exposed Part 2 population. The registry does not report the detail needed to assess event types, timing, or causal attribution.

In general, a challenge study can characterize short-term biological responses under the specific study conditions, but its findings do not automatically generalize to other populations, doses, settings, or clinical outcomes. Here, the numerical biomarker results, analysis methods, and full missing-data details are not reported in the registry record. Conclusions should therefore be confined to the study's stated design and the descriptive information available, without inferring efficacy or clinical implications.

13. Why This Trial Matters Statistically

This trial raises statistical issues common to exploratory biomarker studies. Its central measurement is change from a pre-challenge baseline, repeated over several scheduled timepoints. That design can characterize response timing as well as average change, but its interpretation depends on how within-participant correlation, incomplete measurements, and the multiplicity of outcomes are handled. The registry record defines the measurement and summary statistics but does not report the analysis model or numerical estimates.

The study also combines two challenge agents, dose exploration, and multiple assessment settings, including blood and cantharidin-induced blisters. These features can make a study scientifically informative while complicating a simple single-comparison interpretation. Clear specification of which outcomes and timepoints are primary, and how multiple comparisons are handled, is important when many biomarkers are assessed. The registry lists 36 outcome measures, but does not describe a multiplicity strategy or a formal comparison that would anchor interpretation.

Randomization is relevant to how treatment groups are formed, while open-label conduct and the small enrollment shape how results should be interpreted. Randomization does not eliminate uncertainty in a small exploratory study, and a parallel design does not by itself ensure that every outcome supports a clean comparison between agents. Details of the randomization process, masking of laboratory assessments, and sample-size justification are not reported in the registry record.

Finally, the registry illustrates why outcome definitions and numerical results must be read together. A precise definition of baseline and sampling schedule is useful, but without the observed means, variability, and inferential estimates, it is not possible to assess the magnitude or precision of the response. The statistical value of the record is therefore primarily in its description of the measurement framework and its limits, not in a quantitative conclusion about biomarker effects.

14. 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 for interpreting the posted results is that numerical estimates for the primary endpoint are not reported in the registry record. The record names the biomarkers, units, baseline rule, timepoints, and summary measures, but does not provide the corresponding mean changes, standard deviations, confidence intervals, or tests. It also lists 0 analyses. The extent to which the planned measurements produced usable results cannot be assessed from these details alone.

The study is exploratory and has enrollment of 12 across dose groups and study parts. The registry does not provide the number analyzed for each biomarker at each timepoint, so the precision of individual summaries cannot be assessed. Its available-data approach could result in different analysis counts across visits; reasons and patterns of missingness are not reported. There is no reported missing-data sensitivity analysis.

Several design and analysis details are absent. The randomization method, allocation concealment, laboratory masking, statistical model for repeated measurements, multiplicity approach, interim analysis procedures, and sample-size justification are not reported in the registry record. The record lists many outcome measures, but does not state a hierarchy for interpreting them. The study description is also incomplete at its end, so the remaining details of the continuation plan are not reported.

The registry's condition listing of Arthritis, Rheumatoid does not match its description of healthy participants, and the record does not explain the discrepancy. In addition, the GM-CSF arm dose range is expressed in µg/kg, while the Part 1 adverse-event category is labeled in µg/m^2; the record does not clarify the difference. These reporting limitations warrant care when connecting the safety categories with the described dosing regimen.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Lipopolysaccharide (LPS) or Granulocyte-Macrophage Colony-Stimulating Factor (GM-CSF) Challenge Study on Healthy Subjects — NCT03306589.
• Registry record last updated 2019-08-08; status last verified 2019-06.
• Results first posted 2019-08-08.
• 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.

18. Record Summary

NCT03306589 is a completed, randomized, open-label, parallel-group Phase 1 study of LPS and GM-CSF challenge agents in healthy participants. The registry lists enrollment of 12 and describes dose exploration followed by a continuation part. Cantharidin-induced blisters, blood, and urine were included in the broader biomarker assessment plan.

The listed primary outcome is Part 1 change from baseline in blood TNF-alpha and IL-6 in the LPS arm, summarized by mean and standard deviation at specified timepoints. The registry record defines baseline and the Safety Population but does not report numerical primary outcome estimates or inferential analyses. Part 1 serious adverse event counts are zero in the listed groups; Part 2 categories report zero at risk. These safety figures are descriptive and do not establish comparative safety.