An Early Bactericidal Activity, Safety and Tolerability of GSK3036656 in Subjects With Drug-sensitive Pulmonary Tuberculosis: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for An Early Bactericidal Activity, Safety and Tolerability of GSK3036656 in Subjects With Drug-sensitive Pulmonary Tuberculosis (NCT03557281), a randomized phase 2 study. The primary outcome is Change in log10 Colony Forming Units (CFU) Per (/) Milliliter (mL) of Direct Respiratory Sputum Samples From Baseline to Day 14.

2. Clinical Question

The endpoint reflects early bactericidal activity: it measures change in the quantity of bacteria recovered from sputum during the treatment period. It is an early, laboratory-based outcome rather than a measure of long-term clinical outcomes. The trial record describes GSK3036656 as a compound being investigated for tuberculosis and identifies the study as a dose-escalation trial. The posted results therefore address short-term changes in sputum bacterial counts across dose cohorts, rather than establishing effects on outcomes beyond the study period.

3. Trial Design

This was a Phase IIa, randomized, open-label trial with a sequential model. The registry record lists an enrollment of 76 participants and describes up to four sequential cohorts. Participants within each cohort were randomized in a 3:1 ratio to GSK3036656 or standard-of-care Rifafour e-275. Masking was none, so the trial was open-label. The record is marked completed, with a start date of 2019-03-22 and primary completion date of 2021-12-03.

The GSK3036656 treatment groups had maintenance doses of 1 mg, 5 mg, 15 mg, or 30 mg once daily from Day 2 through Day 14. Each dose group also had a Day 1 loading dose: 3 mg, 15 mg, 30 mg, or 75 mg, respectively. The comparator was Rifafour e-275 once daily from Day 1 to Day 14. After the study treatment period, participants were to receive standard tuberculosis treatment. The approximate duration for an individual participant was 5 weeks, including screening, treatment, and final follow-up.

The primary outcome was the change in log10 CFU/mL in direct respiratory sputum samples from baseline to Day 14. The posted primary outcome summaries include 18 participants in the standard-of-care group and between 9 and 17 participants in the GSK3036656 groups. The reported arm counts differ across outcome measures, reflecting the numbers contributing to each measure; reasons for specific differences are not reported in the registry record.

The registry record does not report the trial’s original sample-size justification, the intended statistical analysis plan, or the rationale for the dose sequence. It lists no reported analyses. The numerical comparisons discussed in the posted results are unadjusted calculations from the arm summaries, rather than a reported sponsor analysis.

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: A total of 76 participants were enrolled (Enrolled Population consisted of all participants who passed screening and entered the study) in this study.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Efficacy Population consisted of participants in the safety population who provided at least two evaluable overnight sputum samples.

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 Population consisted of all randomized participants who received at least one dose of study treatment.

6. Primary Endpoint

The registry lists the primary outcome as Change in log10 Colony Forming Units (CFU) Per (/) Milliliter (mL) of Direct Respiratory Sputum Samples From Baseline to Day 14, measured over Baseline and up to Day 14.

The primary endpoint measures the mean change in log10 CFU/mL in direct respiratory sputum samples from baseline to Day 14. CFU refers to colony forming units, a measure based on bacterial colonies recovered from a sample. Log10 expresses the count on a base-10 logarithmic scale, while CFU/mL represents the amount of bacteria per milliliter of sputum. The registry record defines baseline as the mean of Day -2 and Day -1 measurements; if only one of those time points had data, that value was used as baseline.

The endpoint is used to assess early bactericidal activity because serial sputum measurements can describe whether bacterial counts changed during the 14-day treatment period. A negative mean change indicates a decrease in the measured log10 CFU/mL from baseline; the size and precision of the change describe how much evidence the posted data provide about that early shift. The endpoint is not, by itself, a measure of longer-term tuberculosis outcomes or a full assessment of clinical benefit.

The registry record describes how log(CFU) was calculated from bacterial counts on sputum culture plates, including adjustments for specimen dilution and inoculation volume. The primary outcome used the efficacy population, defined as participants in the safety population with at least two evaluable overnight sputum samples. This restriction means the primary summary is based on participants meeting that sputum-sample criterion, not necessarily everyone enrolled.

7. Results Posted in the Registry

• GSK3036656 5 mg: mean -0.09 versus -0.20 on control; difference 0.11 (95% CI 0.07 to 0.14), standardized effect d = 1.92; z-test p = <0.001.
• GSK3036656 15 mg: mean -0.09 versus -0.20 on control; difference 0.11 (95% CI 0.07 to 0.14), standardized effect d = 1.91; z-test p = <0.001.
• GSK3036656 30 mg: mean -0.14 versus -0.20 on control; difference 0.06 (95% CI 0.02 to 0.10), standardized effect d = 1.09; z-test p = 0.002.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Change in log10 CFU/mL of Direct Respiratory Sputum Samples From Baseline to Day 2

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard errors.
• Result:
  • GSK3036656 5 mg: mean -0.15 versus -0.60 on control; difference 0.44 (95% CI 0.22 to 0.67), standardized effect d = 1.35; z-test p = <0.001.
  • GSK3036656 15 mg: mean -0.06 versus -0.60 on control; difference 0.54 (95% CI 0.32 to 0.77), standardized effect d = 1.64; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for GSK3036656 5 mg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for GSK3036656 15 mg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Change in log10 CFU Per mL of Direct Respiratory Sputum Samples From Day 2 to Day 14

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard errors.
• Result:
  • GSK3036656 5 mg: mean -0.08 versus -0.16 on control; difference 0.08 (95% CI 0.03 to 0.13), standardized effect d = 1.18; z-test p = <0.001.
  • GSK3036656 15 mg: mean -0.09 versus -0.16 on control; difference 0.08 (95% CI 0.03 to 0.12), standardized effect d = 1.07; z-test p = 0.002.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for GSK3036656 5 mg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for GSK3036656 15 mg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Change in log10 Time to Sputum Culture Positivity (TTP) From Baseline to Day 14

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard errors.
• Result:
  • GSK3036656 5 mg: mean 0.01 versus 0.03 on control; difference -0.01 (95% CI -0.01 to -0.01), standardized effect d = -2.28; z-test p = <0.001.
  • GSK3036656 15 mg: mean 0.02 versus 0.03 on control; difference -0.00 (95% CI -0.01 to -0.00), standardized effect d = -0.89; z-test p = 0.010.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for GSK3036656 5 mg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for GSK3036656 15 mg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The secondary outcomes cover additional sputum bacterial-count changes over shorter and later intervals, sputum culture time to positivity, and plasma drug concentration measures. The posted results include summaries for changes in CFU/mL from baseline to Day 2 and from Day 2 to Day 14, as well as changes in log10 time to sputum culture positivity. Pharmacokinetic outcomes are listed, but numerical results for those measures are not included in the results.

At a broad level, the secondary sputum measures add context to the primary baseline-to-Day 14 endpoint by examining when changes were observed and by using a culture-positivity measure alongside CFU counts. The pattern is not uniform across every measure or dose. This is one reason to treat secondary outcomes as complementary evidence rather than as independent confirmation of the primary result. They address related but nonidentical measurements, intervals, and group summaries.

Unadjusted recalculations are posted for selected comparisons in the secondary CFU and time-to-positivity measures. These are not reported sponsor analyses, and no multiplicity correction is reported. When many outcomes, time intervals, or doses are examined, some nominally small p-values can occur by chance even if there is no corresponding underlying effect. A multiplicity plan can address that issue, but the registry record does not report whether one was specified or applied.

The secondary results therefore help describe the short-term outcome pattern, but they carry less confirmatory weight than a prespecified primary analysis with a reported statistical plan. The record does not report analyses for all listed secondary measures, and the presence of an outcome in the registry does not mean that a comparison or numerical result was posted for it. Conclusions should remain limited to the results actually reported and the short treatment period.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Change in log10 CFU/mL of Direct Respiratory Sputum Samples From Baseline to Day 2
• Change in log10 CFU Per mL of Direct Respiratory Sputum Samples From Day 2 to Day 14
• Change in log10 Time to Sputum Culture Positivity (TTP) From Baseline to Day 14
• Change in log10 TTP From Baseline to Day 2
• Change in log10 TTP From Day 2 to Day 14
• Area Under the Plasma Drug Concentration Versus Time Curve From Time Zero to Last Time of Quantifiable Concentration (AUC[0-t]) Following Once Daily Dosing of GSK3036656
• Area Under the Plasma Concentration Time Curve From Zero to 24 Hours (AUC[0-24]) Following Once Daily Dosing of GSK3036656
• Maximum Observed Plasma Drug Concentration (Cmax) Following Once Daily Dosing of GSK3036656

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is continuous: each participant’s outcome is a change in log10 CFU/mL, and the arm summaries report means. A mean describes the average observed change among participants contributing to that arm’s outcome. It does not show the full distribution of individual responses, so the mean alone cannot reveal whether changes were similar across participants or varied substantially.

The arm summaries also report standard errors. In general, a standard error describes the precision of an estimated mean: smaller standard errors indicate that the mean is estimated more precisely, while larger standard errors indicate more sampling uncertainty. A standard error is not the same as the spread of individual participant outcomes. Its interpretation is tied to the mean estimate and the number and variability of observations contributing to it.

The registry record defines baseline as the mean of the Day -2 and Day -1 values when both were available, or the available value when only one time point had data. The primary measure then reflects change from that baseline through Day 14. The efficacy population consisted of participants in the safety population who provided at least two evaluable overnight sputum samples. This population definition is important: the result describes participants meeting the stated measurement requirement, and the record does not report a separate intention-to-treat estimate for all randomized participants.

For three GSK3036656 doses, the posted comparisons give mean differences against standard-of-care, confidence intervals, and unadjusted p-values. These are simple unadjusted recalculations from posted summary statistics. A mean difference expresses the distance between two arm means on the outcome scale. The confidence interval describes a range of values compatible with the estimated difference under the calculation’s assumptions. If a confidence interval includes zero, the data are compatible with no mean difference as well as with differences in either direction; if it lies entirely on one side, it indicates a difference in that direction under the same assumptions. The reported intervals for the three recalculated primary comparisons lie above zero.

A p-value measures how unusual results at least as incompatible with a specified null hypothesis would be under that hypothesis and the calculation’s assumptions. It is not the probability that the null hypothesis is true, nor does it measure clinical importance. The p-values for these comparisons are unadjusted, so they do not account for examining multiple doses or outcomes. In addition, no sponsor-reported analysis or statistical analysis plan is listed in the registry record. The sponsor’s own analysis would take precedence if it were available; the registry record does not report one.

The comparison estimates depend on the posted means and standard errors and on the unadjusted approach used to derive them. They do not include information from participant-level data that could support alternative adjustment, covariate modeling, or checks of analysis assumptions. Such methods are general possibilities, not analyses reported for this trial. The registry record does not report whether the primary analysis adjusted for baseline values or other covariates.

The endpoint’s logarithmic scale also matters for interpretation. A change in log10 CFU/mL is a change on a transformed scale; it should not be read as a raw difference in colony counts. Log transformation can make multiplicative changes easier to represent and can reduce the influence of highly skewed counts in some settings, but the registry record does not describe diagnostic checks or the rationale for transformation beyond the stated endpoint calculation. The results should therefore be interpreted in the units reported.

The listed outcome uses a mean and standard error, while the comparison calculations provide a pooled standard deviation as part of the unadjusted recalculation. Standard deviation describes variability among observations; standard error describes uncertainty in a mean. These measures answer different questions. The comparison confidence intervals combine information about uncertainty in the group means, whereas the arm standard errors quantify precision for each arm mean.

General statistical context: confidence intervals and p-values are affected by both the estimated difference and the amount of uncertainty in the data. A numerically small p-value does not establish that an effect is large or important, and a confidence interval does not account for bias from open-label conduct, missing observations, or choices made after examining results. Those issues require design and analysis information in addition to the numerical summaries.

The registry record lists 26 outcome measures and 25 posted secondary outcomes, but no reported analyses. It does not report a multiplicity strategy, handling rules for missing outcome data, interim analyses, or a detailed analysis plan. Without those details, it is not possible to determine whether the posted comparisons reflect all planned analyses or how decisions about analysis populations and missing data were made. This uncertainty is separate from the numerical uncertainty represented by a confidence interval.

Why does over-dispersion matter?

In general, over-dispersion means that observed variability is greater than a statistical model expects. If a model does not account for that extra variability, uncertainty may be understated and confidence intervals may be too narrow. The registry record does not report whether over-dispersion was assessed for the primary endpoint, and the posted comparison calculations are based on continuous-outcome summaries rather than a reported count model.

What does randomization contribute?

In general, randomization helps make treatment groups comparable on average, including for characteristics not measured or reported. It does not guarantee balance in a small trial, and it does not eliminate uncertainty from missing outcome measurements or open-label conduct. The registry record describes randomized allocation, but does not report group-level baseline characteristics in the outcome information.

11. Confidence Intervals and Precision

For the primary comparisons with posted recalculations, the estimated mean differences are accompanied by 95% confidence intervals. For GSK3036656 5 mg versus standard-of-care, the estimated difference was 0.107, with a confidence interval from 0.070 to 0.144. For 15 mg, the estimate was 0.107, with a confidence interval from 0.069 to 0.145. For 30 mg, the estimate was 0.061, with a confidence interval from 0.022 to 0.100. Each interval is entirely above zero, so the recalculated estimates are compatible with positive mean differences under the methods used.

The widths of the intervals show that the estimates are not exact. The 30 mg estimate is smaller than the estimates for 5 mg and 15 mg, and its interval also lies at lower values. These intervals describe sampling uncertainty under the calculation assumptions; they do not account for all potential uncertainty from open-label conduct, missing data, multiplicity, or unreported analysis choices. Nor do they establish the importance of a difference for clinical outcomes.

The confidence intervals are for mean differences, not for each group’s individual responses. Arm means and standard errors provide context for the group summaries, but the registry record does not post confidence intervals for the individual arm means. The sample sizes contributing to the primary measure were 18 for standard-of-care and 17, 16, and 15 for the 5 mg, 15 mg, and 30 mg groups, respectively. These group sizes are modest, so precision depends on the observed variability as well as the number of participants.

General explanation: a confidence interval is a range generated by a statistical procedure that, over repeated similar studies under its assumptions, would contain the underlying quantity at the stated rate. It is not a range containing a fixed proportion of participants, and it is not a guarantee that the true value lies inside a particular interval. For this trial, the reported intervals should be read as uncertainty around unadjusted recalculated differences rather than as definitive limits on a sponsor-reported treatment effect.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The planning figures are illustrations based on the observed effect, not the original trial’s design calculations. The assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is set to the stated fraction of the observed effect. For these continuous outcomes, the figures refer to participants per arm. The registry record’s note about total event counts applies to time-to-event planning; these primary comparisons are not time-to-event outcomes.

The 80% power column gives the estimated number of participants per arm for a design that would have an 80% probability of detecting the specified effect if that effect were real, under the planning assumptions. The 90% power column uses a higher probability of detection and therefore generally calls for more participants per arm for the same effect. The rows labeled 100%, 75%, and 50% of the observed effect illustrate how the planning estimate changes when the assumed effect is set at different fractions of the observed estimate.

Smaller effects require more participants to detect with the same power and significance level, because a smaller difference is harder to distinguish from random variation. The figures are not observed power: observed results and a sample size cannot be used to tell whether the trial had a meaningful probability of success after the fact. They are hypothetical planning illustrations using the observed estimates and stated assumptions.

The registry record does not report the original sample-size justification or the trial’s design assumptions. These planning illustrations should therefore be kept separate from the actual enrollment of 76 and from the smaller numbers contributing to individual outcome summaries. They do not establish what sample size was planned for the trial or why that enrollment target was selected.

13. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = 0.107; pooled standard deviation σ = 0.056.
2. n = 2 × (1.96 + 0.842)² × 0.056² / 0.107² = 4.2, rounded up to 5 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
delta <- 0.107                      # observed difference in means
sigma <- 0.0556022454310048                     # pooled standard deviation
n_per_arm <- function(power, frac = 1) {
  ceiling(round(2 * (z_a + qnorm(power))^2 * sigma^2 / (delta * frac)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (t-based, so very slightly larger)
cat("power.t.test:", round(power.t.test(delta = delta, sd = sigma, power = 0.80)$n, 1), "\n")

# Simulation: exact two-sample t-test using simulated means and variances
set.seed(3557281)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 5 (80% power), 6 (90% power)
 75% of observed effect: n per arm = 8 (80% power), 11 (90% power)
 50% of observed effect: n per arm = 17 (80% power), 23 (90% power)
power.t.test: 5.4 
Simulated power at n = 5 per arm: 0.761 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

14. Multiplicity, Missing Data and Analysis Populations

The registry record lists 26 outcome measures, including the primary endpoint and 25 secondary outcomes, but lists no reported analyses. There are multiple dose groups and several secondary measurements and time intervals. Testing or comparing many outcomes can increase the chance of obtaining apparently notable results by chance. A prespecified multiplicity strategy can control how results are interpreted across that set of tests, but the registry record does not report whether such a strategy was planned or used.

The p-values accompanying recalculated comparisons are unadjusted. They therefore describe each comparison on its own and do not account for the full set of dose and outcome comparisons. The record does not report a hierarchy of hypotheses, a family-wise error approach, or another method for handling multiple comparisons. The numerical comparisons should be treated as exploratory signals rather than as a complete confirmatory testing framework.

The primary efficacy population required at least two evaluable overnight sputum samples and membership in the safety population. That definition is reported, but the registry record does not report a detailed account of missing samples, exclusions, or reasons that participants did not contribute to particular outcomes. It also does not report an imputation method or sensitivity analyses for missing data. The arm counts differ among posted outcome summaries, and the reasons for those differences are not reported in the registry record.

Baseline values were defined using Day -2 and Day -1 observations, using a single available value if only one was present. Beyond this rule, details about missing baseline measurements or the handling of missing post-baseline values are not reported in the registry record. The record also does not report interim analyses, an original statistical analysis plan, or a sponsor-reported analysis. These omissions limit assessment of whether the posted population and calculations correspond to all planned analyses.

15. 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

The registry record lists zero participants with serious adverse events in every group: Rifafour e-275, GSK3036656 1 mg, 5 mg, 15 mg, and 30 mg. The numbers at risk were 18, 9, 17, 16, and 15, respectively. No deaths are listed in any group. The record also reports participants with other, non-serious adverse events: 14 for standard-of-care, 7 for 1 mg, 14 for 5 mg, 15 for 15 mg, and 9 for 30 mg.

These counts describe the posted safety results during the trial, not the likelihood of an event in other settings or over longer periods. The groups are small, and the record does not provide event details for the other adverse events in the information. Zero recorded serious adverse events does not establish that serious events cannot occur. Comparisons between groups also need caution because the trial was open-label and the number of participants in each group differed.

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

The posted primary results describe mean changes in log10 CFU/mL from baseline to Day 14 in participants with the required evaluable sputum samples. All reported arm means were negative. For the three doses with unadjusted recalculations, the estimated mean differences versus standard-of-care were positive and their confidence intervals were entirely above zero. These observations describe the measured early sputum-count outcome and its estimated uncertainty.

They do not establish long-term effects, outcomes after the reported treatment period, or effects in people who differ from the trial population. The endpoint is a short-term laboratory measure, not a direct measure of recovery, relapse, transmission, or other clinical outcomes. Those outcomes are not reported in the results. The registry record also does not report a formal dose-response analysis, so the arm pattern should not be treated as proof of a dose-response relationship.

The unadjusted comparisons are recalculations from arm summaries, not sponsor-reported adjusted analyses. Their confidence intervals and p-values do not address the full uncertainty created by multiple comparisons or missing observations. The open-label design, modest group sizes, and absence of a reported statistical analysis plan further constrain interpretation. A result that is statistically distinguishable from zero under an unadjusted calculation is not automatically clinically important.

The posted safety counts show no serious adverse events or deaths in the listed groups, while other adverse events were recorded. These findings characterize the trial’s reported experience only. They do not establish comparative safety across larger populations or longer follow-up. Overall, the registry results support a description of short-term bacteriological and safety observations in this study, while leaving broader efficacy and safety questions unresolved.

17. Why This Trial Matters Statistically

This trial is statistically notable because it evaluates a continuous, microbiological endpoint in a small, sequential dose-escalation study. The primary measure uses serial sputum bacterial counts over a defined treatment period, allowing the study to describe early changes rather than waiting for longer-term outcomes. The four GSK3036656 maintenance-dose groups also make the pattern across doses relevant, while the standard-of-care group provides a contemporaneous comparator.

Randomization offers a basis for comparing treatment groups, but the group sizes contributing to the primary outcome are limited. Precision therefore depends substantially on the observed variability, and estimates from small groups can be unstable. The open-label design is another important feature: while the primary endpoint is laboratory-based, knowledge of treatment assignment can affect aspects of trial conduct or data collection. The registry record does not describe safeguards or blinding of laboratory assessment.

The distinction between the primary outcome and the many secondary outcomes is also important. The record lists numerous measures, but no reported analyses or multiplicity strategy. The unadjusted comparisons are useful descriptive calculations, yet they cannot stand in for a prespecified analysis that addresses all planned hypotheses. Likewise, planning illustrations based on the observed effects are not evidence that the trial had adequate power when designed.

For biostatisticians and clinical researchers, this record demonstrates why interpretation requires more than a p-value. Outcome definition, analysis population, arm size, precision, multiplicity, missing-data handling, and study design all shape what can be concluded. The posted data are informative about early sputum changes, but the unreported analysis details limit confirmatory interpretation.

18. 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 trial enrolled 76 participants overall, but the primary endpoint summaries include smaller groups, with 9 to 18 participants per arm. The registry record does not explain all differences in outcome-specific counts. This limits how much can be learned about individual variability and makes estimates vulnerable to imprecision.

The design was open-label and used sequential cohorts. The record does not report whether sputum laboratory assessments were blinded to treatment assignment, how allocation was concealed, or the rationale for cohort sequence and dose selection. These details matter when assessing the possibility of bias and the comparability of dose groups.

The primary outcome covers baseline to Day 14, and the individual participant study duration was approximately 5 weeks. This is a short-term assessment. The registry record does not report longer-term microbiological or clinical outcomes in the results. It also does not report a formal dose-response analysis or a numerical primary comparison for the 1 mg group.

Further limitations are the absence of a reported statistical analysis plan, sponsor-reported analyses, multiplicity strategy, detailed missing-data handling, and original sample-size justification. The unadjusted recalculations use posted summary statistics and cannot replace analyses of participant-level data. The results should therefore be read as descriptive evidence about early bactericidal activity in the measured population, with meaningful uncertainty about the broader treatment effects and the analysis choices.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: An Early Bactericidal Activity, Safety and Tolerability of GSK3036656 in Subjects With Drug-sensitive Pulmonary Tuberculosis — NCT03557281.
• Registry record last updated 2023-10-26; status last verified 2023-10.
• Results first posted 2023-10-26.
• 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.

22. Record Summary

NCT03557281 was a completed, randomized, open-label Phase IIa trial of GSK3036656 in participants with drug-sensitive pulmonary tuberculosis. The study used sequential dose cohorts and compared GSK3036656 with standard-of-care Rifafour e-275. Its primary endpoint was the mean change in log10 CFU/mL of direct respiratory sputum samples from baseline to Day 14.

All posted primary arm means were negative. Unadjusted recalculations for the 5 mg, 15 mg, and 30 mg groups estimated positive mean differences versus standard-of-care, with confidence intervals above zero. The registry record lists no serious adverse events or deaths in the treatment groups, while other adverse events were reported. The results describe short-term sputum-count and safety observations; the record does not report the original sample-size justification, a full statistical analysis plan, or multiplicity and missing-data procedures.