A Phase I, 2-part (Part 1 Being a Single Dose Escalation and Part 2, a Parallel Group) Study of Toll-like Receptor (TLR4) Agonist (GSK1795091) in Healthy Subjects: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Phase I, 2-part (Part 1 Being a Single Dose Escalation and Part 2, a Parallel Group) Study of Toll-like Receptor (TLR4) Agonist (GSK1795091) in Healthy Subjects (NCT02798978), a randomized phase 1 study. The primary outcome is Number of Participants With Non-serious Adverse Events (AE) and Serious Adverse Events (SAE).

2. Clinical Question

3. Trial Design

This was a randomized Phase I study of GSK1795091 in healthy subjects, with an overall registry enrollment of 42. Its title and summary describe two parts with different purposes and procedures. Part 1 was a single-dose escalation study: sequential cohorts received intravenous GSK1795091 or matching placebo, starting at 7 ng, with escalation continuing until the highest dose was evaluated. The registered Part 1 design was randomized, double-blind, sponsor-unblinded, and placebo-controlled.

Part 2 was an open-label, parallel-group evaluation of two doses of GSK1795091. The first dose was administered on Day 1 at a dose determined by the Part 1 results. Cohort 1 received its second dose on Day 8, while Cohort 2 received its second dose on Day 15. The registry’s general design fields identify randomization, double masking, and a parallel model, but the detailed study description distinguishes the open-label Part 2 from the blinded Part 1. The registry record does not report further details explaining how those overall design fields apply across the two parts.

The study began on 2017-01-10, had a primary completion date of 2017-10-13, and is listed as completed. The planned total duration was approximately 10 weeks from screening to the last study visit. Its stated goals included assessing safety, tolerability, pharmacodynamics, and pharmacokinetics. The primary endpoint discussed in the registry is a safety count through Day 32; the listed secondary outcomes include pharmacokinetic measures and changes in IL-6, TNF-alpha, and IFN-gamma concentrations for Part 1. A detailed statistical analysis plan is 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: Forty two participants were randomized in Part 1 of which 2 did not receive dose due to abnormal findings during pre-dose examination. Participants were randomized in ratio of 3:1 to receive GSK1795091 or matching placebo.

Only the first 6 of 7 groups are shown.

Part 1 (30 Days)

Part 2 (45 Days)

Analysis population for the primary outcome, as defined in the registry: All Subjects Population.

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 was discontinued by the Sponsor prior to its scheduled start.

Only the first 6 of 8 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as Number of Participants With Non-serious Adverse Events (AE) and Serious Adverse Events (SAE), measured over Up to Day 32.

The primary endpoint counts participants with non-serious adverse events and serious adverse events through Day 32. An adverse event is defined as an untoward medical occurrence temporally associated with use of a study medicinal product, whether or not considered related to that product. The registry’s serious-event definition includes events such as death, a life-threatening event, hospitalization or prolonged hospitalization, disability or incapacity, congenital anomaly or birth defect, and events considered serious under medical or scientific judgment. It also includes events associated with liver injury and impaired liver function.

The record states that non-serious adverse events were reported using a 5 percentage threshold, alongside serious adverse events. The endpoint therefore summarizes safety events as participant counts rather than measuring a change in a laboratory value or a pharmacokinetic concentration. Counting participants with events is useful in an early-phase study because it can describe the occurrence of reported harms across the dose groups. It does not, by itself, establish whether an event was caused by treatment, how severe non-serious events were, or whether differences between groups are statistically reliable.

The endpoint population is defined as participants enrolled who received a dose of GSK1795091 or placebo. The registry lists the All Subjects Population. It does not report event-level detail, timing within the follow-up period, or a separate analysis for participants who did not receive study medication.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

General statistical explanation: a binary safety endpoint records whether each participant experienced an event or did not experience an event during the specified follow-up. The registry presents counts of participants with events and counts at risk for each group. Such counts make the observed data easy to describe, but they do not by themselves quantify uncertainty or test a difference between treatments. A statistical comparison would require an analysis method and its assumptions, neither of which is reported for this endpoint.

What do the counts represent?

The outcome is a participant count, not a count of every adverse-event episode. A participant who had more than one event may still be represented as one participant with an event for the relevant category. The registry describes separate reporting of non-serious adverse events and serious adverse events. It does not provide participant-level records or explain whether any participants experienced both categories, so the counts should not be combined or treated as mutually exclusive categories.

The non-serious adverse-event reporting threshold is stated as 5 percentage. This means the posted endpoint does not necessarily represent every non-serious event observed in the study; events below the stated reporting threshold may not appear in the reported count. The registry record does not provide the event terms or clarify the detailed application of that threshold across groups. It does define serious adverse events separately using clinical seriousness criteria.

Why are small groups hard to compare?

General statistical explanation: when groups are small, each participant’s outcome has a substantial influence on the observed count. A difference between observed counts can therefore reflect random variation, a real treatment-related difference, or a combination of both. Confidence intervals are one way to communicate uncertainty around an estimate: they give a range of values compatible with the data under the method used. No confidence intervals or other uncertainty estimates are reported for these adverse-event counts.

For rare outcomes, methods based on large-sample approximations may be unreliable when event counts are sparse. Exact methods or other approaches designed for sparse binary data may sometimes be considered, but the registry record does not report which, if any, method was used. It also does not report a statistical test for the safety endpoint. No new comparison or uncertainty estimate should be inferred from the displayed counts alone.

How should the study structure affect interpretation?

Part 1 combines placebo and active-dose cohorts in a dose-escalation design. The listed groups differ by dose and, for the repeated 60 ng group, by repeat dosing. A comparison across those groups is not automatically equivalent to a single randomized comparison of one intervention against one control. The record does not report the allocation details within each cohort, the analysis model, or an adjustment for cohort structure. Part 2 followed a separate open-label parallel-group description and does not have an interpretable event denominator in the posted outcome entry.

General statistical explanation: over-dispersion is variation greater than a model expects, often relevant when analyzing event rates or clustered data. This endpoint is reported as counts of participants with binary outcomes, and no dispersion parameter or event-rate model is reported. There is therefore no basis in the registry record for claiming that over-dispersion was assessed or that a particular model addressed it.

What can be said about sample size and power?

The registry enrollment is 42, but that number is not a statistical justification for detecting a particular safety difference. The record does not report a sample-size calculation, target effect, or power rationale for this endpoint. General statistical explanation: a planning calculation estimates the sample size needed under stated assumptions; it is not evidence that a trial had a particular probability of detecting an effect after the data are observed. The registry record does not report an observed-power analysis or a prespecified statistical plan for the posted adverse-event counts.

The record lists 54 outcome measures and 0 analysis entries. The absence of registered analysis entries does not establish that no analyses were performed; it means that analysis details are not reported in the registry record. The posted safety counts should therefore be read as descriptive registry results unless an analysis and its methods are separately documented in the record.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists 54 outcome measures, including the primary adverse-event endpoint and secondary pharmacokinetic and pharmacodynamic measures. It records 0 analysis entries. A multiplicity strategy for interpreting results across outcomes, doses, cohorts, or time points is not reported in the registry record. Without such a strategy, the record does not establish how the risk of chance findings across multiple analyses was handled.

Missing-data methods are not reported in the registry record. The primary endpoint population is described as all enrolled participants who received a dose of study medication, and is named the All Subjects Population. The record does not describe how participants with incomplete follow-up, missing event assessments, or other unavailable information were handled. It also does not report whether any such data were missing.

The listed adverse-event counts give at-risk and affected counts for Part 1 groups, while the Part 2 row has zero at risk and zero affected. The record does not explain whether the Part 2 values reflect an absence of collected data, a reporting omission, or another reason. Consequently, Part 2 cannot be interpreted as a zero-event result from these entries.

Analysis populations beyond the All Subjects Population, exclusions from the endpoint analysis, and any sensitivity analyses are not reported in the registry record. These omissions limit assessment of whether the posted counts represent the full set of treated participants and how robust the findings are to alternative assumptions about unavailable data.

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 5% reporting threshold.

Most frequent non-serious adverse event terms

For each Part 1 group, the serious adverse-event count is 0: this applies to placebo and to the listed GSK1795091 dose groups, including repeated 60 ng. The at-risk counts are 10 for placebo and 6 for each listed active group. These are reported counts over the stated endpoint period, up to Day 32. A zero count means that no serious adverse event was reported in that group in this outcome entry; it does not establish that the risk is zero.

The non-serious adverse-event counts differ across the listed groups: 0 for placebo, 2 at 7 ng, 0 at 21 ng, 5 at 60 ng, 3 with repeated 60 ng, and 5 at 100 ng. These figures are participants with reported events, not a measure of event severity or a causal comparison. Non-serious events were reported using the stated 5 percentage threshold, so the entries may not include every such event observed.

Part 2 is listed with 0 at risk and 0 affected. Since the record gives no evaluable at-risk count for Part 2, those values should not be interpreted as confirming that no adverse events occurred. The registry record does not report further Part 2 safety counts in this endpoint.

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

The registry record documents a Phase I study intended to characterize safety, tolerability, pharmacodynamics, and pharmacokinetics of intravenous GSK1795091 in healthy subjects. For the posted primary endpoint, it records no serious adverse events among the participants at risk in the listed Part 1 groups. It also records non-serious adverse events in some Part 1 groups, with counts varying across the doses and the repeated-dose group. This is a descriptive account of what was reported for follow-up through Day 32.

These data do not establish that GSK1795091 is safe or unsafe, nor do they establish that the observed non-serious events were caused by the study drug. The endpoint definition includes events whether or not considered related to the medicinal product. The registry does not provide event-level details or a statistical comparison that would support a causal conclusion. The absence of reported serious events in these groups cannot rule out uncommon or unobserved serious events.

The trial involved healthy subjects, and the registry states that the results would support the design of future studies in people with advanced malignancies receiving GSK1795091 in combination with immune system modulators. The posted safety counts do not answer whether such a future combination is effective or safe in people with cancer. They also do not establish the appropriate dose or dosing schedule for a later study.

The Part 2 safety row has no participants at risk recorded in this endpoint, so it cannot contribute an interpretable event comparison. More broadly, the record does not report confidence intervals, p-values, adjustment methods, or a complete statistical analysis plan. The reasonable interpretation is therefore limited to the recorded Part 1 counts and the design information in the registry record, without extrapolating beyond this study population or endpoint.

13. Why This Trial Matters Statistically

This study illustrates why early-phase safety results need to be read as descriptions of observed events rather than as definitive comparisons. Part 1 used sequential dose escalation, included a placebo group, and reported several dose-specific groups. Those features provide context for observing adverse events during initial evaluation, but they do not turn each group’s count into a precise estimate of risk. The small group sizes mean the posted counts alone cannot characterize uncommon events with certainty.

The distinction between serious and non-serious adverse events also matters statistically. No serious adverse events were reported in the listed Part 1 groups, while non-serious events were reported in some groups. These categories use different criteria, and the registry’s non-serious-event reporting threshold affects what is represented in the posted count. A reader should not treat the categories as interchangeable or assume that the non-serious count is a complete catalogue of all events.

The trial’s two-part structure is another important feature. Part 1 was blinded and placebo-controlled, whereas Part 2 is described as open-label and its posted endpoint row contains no at-risk participants. A broad label for the trial’s design does not eliminate the need to assess each part separately. The record’s lack of analysis details also means that readers cannot evaluate an inferential model, multiplicity approach, or missing-data strategy.

For clinical researchers, the statistical lesson is that transparent denominators, event definitions, and analysis methods are essential to interpreting early-phase safety. In this record, the Part 1 counts offer a limited descriptive view; uncertainty and the absence of Part 2 data constrain what can be concluded. The record does not support claims about efficacy or later clinical outcomes.

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 adverse-event results are reported as participant counts, without confidence intervals, p-values, or adjusted estimates. The registry record does not report a formal statistical comparison between placebo and any GSK1795091 group, or among the active-dose groups. Consequently, the observed differences cannot be interpreted as statistically established effects.

The group sizes are limited, and the record provides no sample-size justification for the safety endpoint. Counts of 0 serious adverse events therefore do not exclude uncommon risks. The registry does not report detailed event terms, severity grades for non-serious events, duration, relationship assessments, or the number of events per participant. Its 5 percentage threshold for reporting non-serious events further limits what can be learned from the posted counts.

The Part 2 row reports 0 participants at risk and 0 affected for both event categories. The reason for this entry is not reported in the registry record, so it cannot be treated as an observed zero-event result. This limits interpretation of the open-label repeat-dose part.

The record describes randomization and double masking at the overall trial level, while its detailed description identifies Part 1 as double-blind and Part 2 as open-label. It does not provide additional allocation or analysis details that resolve how the general design fields apply across parts. Missing-data handling, analysis populations beyond the All Subjects Population, multiplicity control, and the statistical analysis plan are also not reported in the registry record.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: A Phase I, 2-part (Part 1 Being a Single Dose Escalation and Part 2, a Parallel Group) Study of Toll-like Receptor (TLR4) Agonist (GSK1795091) in Healthy Subjects — NCT02798978.
• Registry record last updated 2020-11-27; status last verified 2020-11.
• Results first posted 2019-06-10.
• 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

NCT02798978 is a completed Phase I study of intravenous GSK1795091 in healthy subjects, with placebo included in Part 1. Part 1 evaluated escalating single doses and a repeated 60 ng group under a randomized, double-blind, sponsor-unblinded design. Part 2 was an open-label parallel-group evaluation of two doses administered either 1 week apart or 2 weeks apart.

The primary safety endpoint counted participants with non-serious and serious adverse events through Day 32. The listed Part 1 groups reported 0 serious adverse events; non-serious event counts varied by group. Part 2 has no at-risk participants recorded in this endpoint entry. The registry lists 54 outcome measures and 0 analysis entries. Statistical analysis details, including missing-data and multiplicity methods, are not reported in the registry record.