GKT137831 in IPF Patients With Idiopathic Pulmonary Fibrosis: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for GKT137831 in IPF Patients With Idiopathic Pulmonary Fibrosis (NCT03865927), a randomized phase 2 study. The primary outcome is Surrogate Biomarker of Oxidative Stress by Mass Spectroscopy.

2. Clinical Question

3. Trial Design

This was a phase II, randomized, parallel-group trial comparing GKT137831 with placebo in ambulatory patients with idiopathic pulmonary fibrosis. The registry record lists University of Alabama at Birmingham as the sponsor and describes the trial as multicenter. The recorded enrollment is 58. The study is marked completed, with a start date of 2020-09-07 and primary completion date of 2024-11-30.

The intervention was oral GKT137831, 400 mg twice daily for 24 weeks. The comparator was an identically appearing oral placebo, also taken twice daily for 24 weeks. The registry describes the study as double-blind and records triple masking. It does not specify which participant, investigator, assessor, or other roles were masked, so the meaning of the masking label for each role is not reported in the registry record.

The primary outcome was a continuous biomarker measure: change in circulating o,o'-dityrosine, measured in plasma using mass spectroscopy, from baseline through week 24. The posted summaries report a mean and standard deviation for each arm and a mean-difference estimate with an interval. Outcome-specific group sizes are 21 for placebo and 22 for GKT137831. The record also lists four secondary outcomes: collagen degradation product by enzyme-linked immunoabsorbent assay, pulmonary function by spirometry, six-minute walk distance, and safety assessed by adverse events.

The registry record does not describe the randomization sequence, allocation concealment, the planned primary analysis model, or the trial's original sample-size rationale. These omissions limit conclusions about how the study was designed to address potential bias and how the posted comparison relates to the prespecified analysis plan.

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

Reasons for not completing this period:

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.

6. Primary Endpoint

The registry lists the primary outcome as Surrogate Biomarker of Oxidative Stress by Mass Spectroscopy, measured over From baseline thru week 24.

The primary endpoint concerns o,o'-dityrosine in plasma, measured by mass spectroscopy and reported in micromole/M tyrosine. The registry describes comparing changes in concentration from baseline through week 24, including absolute concentrations and percentages of baseline, within and between the treatment arms. The posted primary outcome summaries give mean changes and standard deviations for the continuous measure.

The biomarker was selected as a surrogate measure of oxidative stress. In the registry's biological rationale, nicotinamide adenine dinucleotide phosphate oxidase enzymes generate reactive oxygen species, which are hypothesized to contribute to pulmonary injury in idiopathic pulmonary fibrosis. Measuring a circulating oxidative-stress marker can therefore address a biological question about changes in that marker during treatment. It does not, by itself, establish a change in symptoms, lung function, walking ability, disease progression, or other outcomes that matter directly to patients.

Those broader clinical domains appear among the listed secondary outcomes, but results for them are not included in the posted summaries described in the registry record. The record also does not explain how the biomarker's absolute and percentage changes were handled in the primary analysis or state a threshold for a clinically important difference. Consequently, interpretation should remain focused on the reported continuous biomarker comparison rather than treating it as a direct measure of clinical benefit.

7. Results Posted in the Registry

• GKT137831: mean 13.30 versus -79.90 on control; difference 93.20 (95% CI -17.30 to 203.70), standardized effect d = 0.52; z-test p = 0.098.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is continuous, and the posted arm summaries use means and standard deviations. A mean is the arithmetic average of measurements in a group. A standard deviation describes how dispersed observations are around that group's mean. In general, a mean difference compares the averages in two groups on the same measurement scale. It retains the outcome's units, which in this case are micromole/M tyrosine.

What does the mean difference represent?

The registry reports a between-group mean-difference estimate of 93.2. This is an estimate of the contrast between the mean changes in the two arms, rather than a count of participants or a percentage. The group summaries are -79.9 for placebo and 13.3 for GKT137831. Because the endpoint is a change from baseline, interpretation depends on how the change was defined and analyzed. The description says absolute concentrations and percentages of baseline were to be compared, but details of the analysis producing the posted estimate are not reported in the registry record.

In general, estimates from samples vary from one study to another, even when the underlying effects are similar. A confidence interval is a way to express the uncertainty around an estimate under a statistical model: it gives a range of values compatible with the data and model at the interval's stated confidence level. The registry record gives interval bounds of -17.3 and 203.7 but does not identify the confidence level in the information. The interval spans zero, so its range includes no difference as well as differences in either direction.

How should the p-value be understood?

The unadjusted p-value is 0.098. In general, a p-value measures how unusual data at least as extreme as those observed would be if a specified null hypothesis and statistical model were correct. It is not the chance that the treatment has no effect, nor does it state the probability that the observed result occurred randomly. A p-value also does not describe the magnitude or precision of the estimate. Here it is explicitly unadjusted; the registry record does not report whether the trial's planned primary analysis used an adjustment, or what decision threshold it specified.

In general, the term unadjusted can matter when several outcomes or comparisons are examined: repeated testing can increase the chance of at least one small p-value under a global null. The registry lists one primary outcome and four secondary outcomes, but the registry record does not report a multiplicity strategy, the number of comparisons in the planned analysis, or whether secondary analyses were confirmatory or exploratory. It would not be appropriate to assume that no adjustment was planned merely because the posted p-value is labeled unadjusted.

Why does variation matter?

The reported standard deviations are 255.5 for placebo and 39.3 for GKT137831. In general, greater dispersion means measurements are more spread out around the group mean. Unequal dispersion can affect how precisely group means are estimated and may matter for the choice or behavior of an analysis method. The registry record does not identify the model, variance assumptions, or any method used to address the different observed standard deviations. The reported pooled standard deviation is 180.7, but its role in the sponsor's primary analysis is not described.

The outcome-specific group counts are 21 and 22. In general, estimates based on limited numbers of observations can be sensitive to individual measurements, especially when variation is substantial. That general consideration is relevant to interpreting a mean and its interval, but it does not establish that any particular observation drove this result. The registry record does not provide participant-level data or diagnostics that would support such an assessment.

What is not documented about the analysis?

The record reports zero analyses in its analysis listing, alongside posted outcome summaries and a comparison. It does not provide the statistical analysis plan, the model specification, covariate adjustments, treatment of repeated measurements, or methods for missing data. It also does not state which analysis population was used, such as all randomized participants or a subset defined by treatment received or available measurements. These details are important for determining exactly what population and estimand the estimate represents.

In general, randomization supports comparisons between assigned groups, while masking can reduce some opportunities for expectations to influence care or outcome assessment. These design features do not replace a clear analysis description, and they do not eliminate uncertainty from missing outcomes or imprecision. The trial is recorded as randomized and triple-masked, but procedural details of randomization and masking are not reported in the registry record.

The sample-size figures are planning illustrations based on the observed effect, using a two-sided significance level of 0.05, equal allocation, and a normal approximation. They are not a description of the trial's original design assumptions. In general, such calculations ask how many participants per arm a future study might require to detect a specified effect under stated assumptions. They do not establish what effect will occur in a new study, and the registry record does not report the original sample-size justification.

10. Confidence Intervals and Precision

The posted mean-difference estimate is 93.2 micromole/M tyrosine, with interval bounds of -17.3 and 203.7. The registry record does not identify the confidence level associated with these bounds, so they are described as the reported interval rather than assigned a confidence level that is not stated. The interval is wide relative to the point estimate and includes zero. Its values also extend to both negative and positive differences, leaving uncertainty about the direction and size of the between-group difference.

An estimate is the single best summary value from the reported comparison, while an interval conveys how uncertain that estimate is under the analysis used. In general, a narrower interval indicates greater precision, and a wider interval indicates less precision. Precision is not the same as clinical importance: a precise estimate could be small or large, and a clinically important effect could still be estimated imprecisely.

The outcome-specific groups contain 21 placebo participants and 22 GKT137831 participants in the posted summaries. The standard deviations differ substantially, at 255.5 and 39.3. In general, the amount of variation and the number of observations both influence the precision of an estimated mean or difference. The registry record does not give enough detail about the analysis model, missing data, or interval construction to assess their effects further.

The interval should not be read as showing that all values inside it are equally likely, nor as a range that must contain the true effect. It is also not evidence by itself that the observed difference is clinically meaningful. The registry record does not report a threshold for a meaningful biomarker change. The posted interval therefore conveys substantial uncertainty in the estimated biomarker contrast without resolving its clinical interpretation.

11. 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 participants per arm, not total events. They are illustrations based on the observed effect, with a two-sided significance level of 0.05, equal allocation between the arms, and a normal approximation. The figures describe planning for 100%, 75%, or 50% of the observed effect. The trial's original sample-size assumptions and justification are not reported in the registry record.

The 80% power column gives the estimated number of participants per arm for a future study to have an 80% chance of detecting the chosen effect under the stated assumptions, if that effect and the planning model were correct. The 90% power column expresses the corresponding planning target for a 90% chance. These are design targets, not guarantees about what a study will achieve. For planning around 100% of the observed effect, the figures are 59 participants per arm for 80% power and 79 for 90% power. For planning around 75% of the observed effect, they are 105 and 141. For planning around 50% of the observed effect, they are 236 and 316.

Smaller effects are harder to distinguish from random variation, so planning to detect a smaller fraction of the observed effect requires more participants per arm in these illustrations. The observed effect comes from a study with limited outcome-specific group counts and an imprecise estimate; a planning exercise based on it may not predict the effect in another study. These figures are not observed power.

12. 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 = 93.200; pooled standard deviation σ = 180.652.
2. n = 2 × (1.96 + 0.842)² × 180.652² / 93.200² = 59.0, rounded up to 59 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 <- 93.2                      # observed difference in means
sigma <- 180.651887989432                     # 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(3865927)
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 = 59 (80% power), 79 (90% power)
 75% of observed effect: n per arm = 105 (80% power), 141 (90% power)
 50% of observed effect: n per arm = 236 (80% power), 316 (90% power)
power.t.test: 60 
Simulated power at n = 59 per arm: 0.794 (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.

13. Multiplicity, Missing Data and Analysis Populations

The registry lists a primary biomarker outcome and four secondary outcomes. The secondary outcomes concern collagen degradation product, pulmonary function by spirometry, six-minute walk distance, and safety assessed by adverse events. The record also gives a count of five outcome measures, but only four secondary outcomes are listed. The statistical relationship between that count and the listed outcomes is not explained in the registry record.

The reported p-value for the primary comparison is labeled unadjusted. The registry record does not report a multiplicity plan, such as how statistical error was to be handled across outcomes, time points, or other comparisons. Without such a plan, the role of the unadjusted p-value in the trial's overall inferential strategy cannot be determined from the registry record. Results for the listed secondary outcomes are not included in the posted outcome summaries, so their estimates and uncertainty cannot be evaluated.

Details about missing data are also not reported in the registry record. It does not state how participants without a week 24 measurement were handled, whether any values were imputed, or whether the analysis used all randomized participants or another population. The record reports enrollment of 58, while the primary outcome summaries report 21 participants for placebo and 22 for GKT137831. These figures describe different entries in the record; it does not explain their relationship. They should not be used to infer why outcome-specific counts differ from enrollment.

The analysis listing reports zero analyses, although an outcome comparison is posted. No statistical analysis plan, prespecified model, covariate adjustment, or analysis-population definition is given in the registry record. The record also does not state whether an interim analysis was planned or performed. These missing details limit the ability to judge whether the posted estimate follows a prespecified method or how robust the result is to different reasonable analytical choices.

In general, multiplicity, missing outcomes, and analysis-population choices can affect the interpretation of estimates and p-values. That general point does not indicate that any particular problem occurred in this trial. The available registry information is insufficient to determine how these issues were addressed.

14. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: acute respiratory dysfunction (16.7%), hospitalization for elective surgery (7.1%), neurologic complications (syncope, lighteadedness) (3.3%), prolongation of prothrombin time (3.3%).

The posted safety summaries give serious adverse events for 7 of 30 participants at risk in the GKT137831 group and 3 of 28 at risk in the placebo group. Other adverse events are reported for 19 of 30 and 17 of 28, respectively. Deaths are recorded as 0 in each group. These counts and denominators describe the safety entries; the registry record does not explain the relationship between these at-risk groups and the outcome-specific groups reported for the biomarker analysis.

The listed serious-event terms include Acute Respiratory Dysfunction at 16.7%, Hospitalization for elective surgery at 7.1%, Neurologic Complications (syncope, lighteadedness) at 3.3%, and Prolongation of Prothrombin Time at 3.3%. The record does not provide further detail in these summaries about event timing, duration, severity beyond the serious-event classification, or whether an individual could contribute more than one event term. These figures should not be treated as a complete account of safety without those details.

In general, adverse-event counts describe events recorded during a study and do not by themselves establish that treatment caused an event. The registry record does not report a comparative safety analysis or provide enough information to assess differences in risk. These summaries describe reported events and denominators only.

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

The result describes a between-group comparison in a circulating oxidative-stress biomarker from baseline through week 24. The posted mean-difference estimate is 93.2 micromole/M tyrosine, with interval bounds of -17.3 and 203.7 and an unadjusted p-value of 0.098. The interval includes zero and extends in both directions, so the reported estimate does not precisely establish the direction or size of the difference in mean biomarker change.

The biomarker was chosen to address a biological hypothesis concerning oxidative stress. A biomarker result is not interchangeable with a clinical outcome. Even a difference in a circulating marker would not, on its own, show a corresponding change in lung function, walking distance, symptoms, disease course, or survival. The registry lists clinical and safety outcomes as secondary measures, but their results are not included in the posted outcome summaries.

The statistical comparison also does not establish the mechanism proposed in the trial summary. The summary describes a rationale involving NOX enzymes and reactive oxygen species, but the posted estimate alone cannot determine whether that mechanism was affected or whether a biomarker change reflects a clinically relevant process. The registry record does not state a threshold for a clinically important biomarker difference.

The estimate is based on outcome-specific group counts of 21 and 22, with different reported standard deviations in the two arms. The record does not give participant-level data, missing-data methods, analysis-population details, or the planned analysis model. This limits assessment of the robustness and generalizability of the result. The registry record also does not provide enough information to connect the biomarker comparison to the full enrolled group.

Accordingly, the posted result is best read as an imprecise estimate of a continuous biomarker contrast in the analyzed groups. It is not a stand-alone demonstration of clinical benefit or harm, and it does not settle the biological or clinical questions motivating the trial.

16. Why This Trial Matters Statistically

This trial illustrates a common statistical challenge in phase II research: a biologically motivated endpoint can be measured quantitatively, while the estimate remains uncertain and its clinical meaning is not self-evident. The primary measure is a continuous circulating biomarker, which permits a mean-difference comparison, but it is a surrogate rather than a direct measure of how participants feel or function.

The posted estimate is accompanied by a broad interval spanning negative and positive values. This makes precision central to interpretation. Reporting only a point estimate or p-value would leave out the uncertainty around the comparison. The different standard deviations across groups also highlight why a reader should consider the spread of measurements alongside the means; a group average can mask substantial variation among observations.

The record is useful for distinguishing the trial's scientific rationale from its reported statistical evidence. The rationale describes a proposed oxidative-stress pathway, while the posted primary comparison quantifies a biomarker change. Neither the hypothesis nor a biomarker result alone establishes a downstream clinical effect. This distinction matters when interpreting studies designed to test an early biological signal.

The record also shows why interpretation depends on transparent reporting beyond the estimate itself. Information about the planned model, missing data, analysis population, multiplicity, and the original sample-size rationale is not reported in the registry record. Those details would help readers understand how the estimate was produced, how secondary results should be interpreted, and whether the analysis followed the intended plan.

Finally, planning illustrations based on the observed effect demonstrate how assumptions about the effect size affect future study requirements. They are useful as conditional design scenarios, not as a verdict on the current study's power or as a forecast of a future result. Statistical interpretation is strongest when these planning calculations are kept separate from the observed trial findings.

17. 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 registry record leaves several trial-specific uncertainties. It reports enrollment of 58, but the primary outcome summaries contain 21 participants in the placebo arm and 22 in the GKT137831 arm. The record does not explain the relationship between those figures, identify the analysis population, or describe the handling of unavailable measurements. The difference should not be attributed to a particular cause without further information.

The primary outcome description refers to both absolute concentrations and percentages of baseline, but the posted summaries give mean changes in the continuous measure. The record does not clarify whether these were separate analyses, how the primary comparison was selected, or how baseline values were incorporated. The statistical model, covariate adjustment, repeated-measure handling, and interval construction are not reported.

The reported standard deviations differ between arms, and the outcome-specific groups are limited in size. This can make the estimate sensitive to variation in the observed data, but the registry record does not provide participant-level results or diagnostics to assess that possibility. It also does not report the original sample-size assumptions, any interim analyses, or a multiplicity strategy. The planning figures based on the observed effect do not substitute for the trial's original design rationale.

The posted summaries do not include results for the listed secondary outcomes. Safety figures are reported, but details about event timing, event-level counts, and the relationship between safety and biomarker analysis populations are not stated. The record also does not specify which roles were masked despite recording triple masking. These limitations affect how fully the trial's design, analysis, and results can be evaluated from the registry record.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: GKT137831 in IPF Patients With Idiopathic Pulmonary Fibrosis — NCT03865927.
• Registry record last updated 2026-05-11; status last verified 2026-04.
• Results first posted 2026-05-11.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

21. Record Summary

NCT03865927 is a completed, randomized, parallel-group phase II trial of GKT137831 versus placebo in ambulatory patients with idiopathic pulmonary fibrosis. It is sponsored by University of Alabama at Birmingham and recorded as multicenter, double-blind, and triple-masked. The study lists enrollment of 58 and treatment over 24 weeks.

The primary outcome was change in plasma o,o'-dityrosine, measured by mass spectroscopy from baseline through week 24. The posted mean-difference estimate is 93.2 micromole/M tyrosine, with interval bounds of -17.3 and 203.7 and an unadjusted p-value of 0.098. The interval includes zero, and the registry record does not report its confidence level or the details of the analysis plan.

Four secondary outcomes are listed, but their results are not included in the posted outcome summaries. The registry record does not report the trial's original sample-size rationale, missing-data methods, or analysis-population definition.